diff --git "a/test_1perm_Falsetemp_Falsetempaug_EgoExOR_5k_samples_drophistory0.5.json" "b/test_1perm_Falsetemp_Falsetempaug_EgoExOR_5k_samples_drophistory0.5.json" new file mode 100644--- /dev/null +++ "b/test_1perm_Falsetemp_Falsetempaug_EgoExOR_5k_samples_drophistory0.5.json" @@ -0,0 +1,148314 @@ +[ + { + "id": "Ultrasound_5_6_2146_0", + "timepoint": 2146, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2146, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,needle,holding; assistant,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,syringe,injecting; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1250_0", + "timepoint": 1250, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1250, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; needle,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2225_0", + "timepoint": 2225, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2225, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; syringe,needle,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2235_0", + "timepoint": 2235, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2235, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,syringe,injecting; head surgeon,needle,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1183_0", + "timepoint": 1183, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1183, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,inserting; assistant,ultrasound probe,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_693_0", + "timepoint": 693, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 693, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,curette,holding; patient,operating table,lyingon; head surgeon,microscope eye,looking; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2470_0", + "timepoint": 2470, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2470, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,patient,manipulating; assistant,patient,manipulating; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_979_0", + "timepoint": 979, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 979, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,instrument table,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_3_6_235_0", + "timepoint": 235, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 235, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; curette,patient,touching; head surgeon,forceps,holding; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,operating table,closeto; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1782_0", + "timepoint": 1782, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1782, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,syringe,holding; needle,patient,touching; head surgeon,needle,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2688_0", + "timepoint": 2688, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2688, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,needle,holding; syringe,needle,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2511_0", + "timepoint": 2511, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2511, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "MISS_1_1_1360_0", + "timepoint": 1360, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1360, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; circulator,operating table,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_113_0", + "timepoint": 113, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 113, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,ultrasound gel,applying; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_232_0", + "timepoint": 232, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 232, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,instrument table,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_419_0", + "timepoint": 419, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 419, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,operating table,closeto; head surgeon,herbal disk,removing; patient,operating table,lyingon; head surgeon,curette,holding; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; head surgeon,forceps,holding; anesthetist,operating table,closeto; assistant,forceps,holding; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_640_0", + "timepoint": 640, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 640, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1322_0", + "timepoint": 1322, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1322, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,inserting; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1311_0", + "timepoint": 1311, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1311, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; body marker,patient,touching; head surgeon,ultrasound probe,controlling; assistant,ultrasound probe,looking; assistant,body marker,applying; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1696_0", + "timepoint": 1696, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1696, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,tissue mark,looking; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1777_0", + "timepoint": 1777, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1777, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; needle,patient,touching; head surgeon,needle,holding; assistant,syringe,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1718_0", + "timepoint": 1718, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1718, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,inserting; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,tissue mark,looking " + } + ] + }, + { + "id": "MISS_2_2_145_0", + "timepoint": 145, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 145, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,scalpel,holding; scalpel,patient,touching; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; anesthetist,operating table,closeto; assistant,instrument table,closeto; head surgeon,patient,cutting; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1033_0", + "timepoint": 1033, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1033, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,microscope,manipulating; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_2120_0", + "timepoint": 2120, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2120, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1388_0", + "timepoint": 1388, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1388, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; head surgeon,microscope controller,holding; anesthetist,operating table,closeto; anesthetist,patient,anaesthetising " + } + ] + }, + { + "id": "Ultrasound_3_4_2391_0", + "timepoint": 2391, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2391, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,needle,holding; patient,operating table,lyingon; syringe,needle,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3176_0", + "timepoint": 3176, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3176, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2194_0", + "timepoint": 2194, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2194, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound screen,looking; syringe,needle,touching; head surgeon,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_1_1_652_0", + "timepoint": 652, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 652, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2449_0", + "timepoint": 2449, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2449, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,needle,holding; patient,operating table,lyingon; needle,patient,touching; syringe,needle,touching; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_456_0", + "timepoint": 456, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 456, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_315_0", + "timepoint": 315, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 315, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; circulator,operating table,looking; patient,operating room,entering " + } + ] + }, + { + "id": "MISS_2_2_555_0", + "timepoint": 555, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 555, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope controller,controlling; anesthetist,health monitor,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_649_0", + "timepoint": 649, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 649, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_604_0", + "timepoint": 604, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 604, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; head surgeon,microscope eye,looking; scalpel,patient,touching; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,scalpel,holding; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; anesthetist,health monitor,looking; head surgeon,curette,holding; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1184_0", + "timepoint": 1184, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1184, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1819_0", + "timepoint": 1819, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1819, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1584_0", + "timepoint": 1584, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1584, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,inserting; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1662_0", + "timepoint": 1662, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1662, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_226_0", + "timepoint": 226, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 226, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,instrument table,looking; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2378_0", + "timepoint": 2378, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2378, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; assistant,needle,removing; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2131_0", + "timepoint": 2131, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2131, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; forceps,cotton,touching; assistant,patient,disinfection; assistant,operating table,closeto; head surgeon,operating table,closeto; cotton,patient,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2198_0", + "timepoint": 2198, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2198, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,needle,holding; assistant,syringe,injecting; syringe,needle,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2119_0", + "timepoint": 2119, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2119, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,needle,holding; syringe,needle,touching; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2698_0", + "timepoint": 2698, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2698, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,syringe,injecting; syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,needle,holding; assistant,needle,holding; needle,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_807_0", + "timepoint": 807, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 807, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,health monitor,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_157_0", + "timepoint": 157, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 157, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,antiseptic,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2108_0", + "timepoint": 2108, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2108, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; syringe,needle,touching; assistant,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_1108_0", + "timepoint": 1108, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1108, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_877_0", + "timepoint": 877, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 877, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1167_0", + "timepoint": 1167, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1167, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_338_0", + "timepoint": 338, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 338, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_920_0", + "timepoint": 920, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 920, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_326_0", + "timepoint": 326, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 326, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_519_0", + "timepoint": 519, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 519, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,curette,holding; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; scalpel,patient,touching; head surgeon,scalpel,holding; head surgeon,herbal disk,removing; curette,patient,touching; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_828_0", + "timepoint": 828, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 828, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3105_0", + "timepoint": 3105, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3105, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,tissue paper,holding " + } + ] + }, + { + "id": "MISS_3_6_222_0", + "timepoint": 222, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 222, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,forceps,holding; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,curette,holding; anesthetist,operating table,closeto; head surgeon,herbal disk,removing; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2861_0", + "timepoint": 2861, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2861, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,syringe,removing; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_195_0", + "timepoint": 195, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 195, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound gel,applying; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_916_0", + "timepoint": 916, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 916, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,body marker,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2538_0", + "timepoint": 2538, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2538, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,patient,manipulating; head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2645_0", + "timepoint": 2645, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2645, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_489_0", + "timepoint": 489, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 489, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,scalpel,holding; anesthetist,operating table,closeto; anesthetist,health monitor,looking; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_169_0", + "timepoint": 169, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 169, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1130_0", + "timepoint": 1130, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1130, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2097_0", + "timepoint": 2097, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2097, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,operating table,closeto; syringe,needle,touching; head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_466_0", + "timepoint": 466, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 466, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,forceps,holding; curette,patient,touching; anesthetist,operating table,closeto; head surgeon,curette,holding; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "MISS_1_1_2408_0", + "timepoint": 2408, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2408, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2492_0", + "timepoint": 2492, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2492, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,needle,holding; head surgeon,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; syringe,needle,touching; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_3_6_385_0", + "timepoint": 385, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 385, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,herbal disk,removing; curette,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; head surgeon,forceps,holding; head surgeon,curette,holding; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1637_0", + "timepoint": 1637, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1637, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,tissue mark,looking; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_513_0", + "timepoint": 513, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 513, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1008_0", + "timepoint": 1008, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1008, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,microscope,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3297_0", + "timepoint": 3297, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3297, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,holding; patient,operating table,lyingon; assistant,bin,looking; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_613_0", + "timepoint": 613, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 613, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; head surgeon,herbal disk,removing; scalpel,patient,touching; head surgeon,scalpel,holding; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,operating table,closeto; curette,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_857_0", + "timepoint": 857, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 857, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; body marker,patient,touching; assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,body marker,applying " + } + ] + }, + { + "id": "MISS_1_1_2431_0", + "timepoint": 2431, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2431, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "MISS_3_6_813_0", + "timepoint": 813, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 813, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_318_0", + "timepoint": 318, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 318, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,operating table,looking; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1707_0", + "timepoint": 1707, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1707, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2752_0", + "timepoint": 2752, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2752, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_783_0", + "timepoint": 783, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 783, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,health monitor,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2274_0", + "timepoint": 2274, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2274, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,syringe,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_824_0", + "timepoint": 824, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 824, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_121_0", + "timepoint": 121, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 121, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2225_0", + "timepoint": 2225, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2225, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; anesthetist,operating table,closeto; assistant,forceps,holding; assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,disinfection; head surgeon,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1138_0", + "timepoint": 1138, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1138, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,body marker,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1131_0", + "timepoint": 1131, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1131, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1920_0", + "timepoint": 1920, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1920, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,applying; assistant,ultrasound probe,looking; assistant,syringe,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_397_0", + "timepoint": 397, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 397, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,herbal disk,removing; curette,patient,touching; head surgeon,curette,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,forceps,holding; anesthetist,operating table,closeto; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_903_0", + "timepoint": 903, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 903, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_299_0", + "timepoint": 299, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 299, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_953_0", + "timepoint": 953, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 953, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,microscope eye,looking; assistant,patient,aspirating; anesthetist,anesthesia equipment,controlling; cotton,patient,touching; assistant,forceps,holding; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1265_0", + "timepoint": 1265, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1265, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; assistant,needle,inserting; assistant,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1401_0", + "timepoint": 1401, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1401, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,needle,removing " + } + ] + }, + { + "id": "Ultrasound_3_4_102_0", + "timepoint": 102, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 102, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon; assistant,patient,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_779_0", + "timepoint": 779, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 779, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,closeto; patient,operating table,lyingon; circulator,microscope,manipulating; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "MISS_1_1_121_0", + "timepoint": 121, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 121, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_92_0", + "timepoint": 92, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 92, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_1_1_9_0", + "timepoint": 9, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 9, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_2144_0", + "timepoint": 2144, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2144, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; assistant,operating table,closeto; assistant,needle,holding; syringe,needle,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1573_0", + "timepoint": 1573, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1573, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1885_0", + "timepoint": 1885, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1885, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1274_0", + "timepoint": 1274, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1274, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,patient,manipulating; anesthetist,anesthesia equipment,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1011_0", + "timepoint": 1011, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1011, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,patient,aspirating; anesthetist,health monitor,looking; head surgeon,microscope eye,looking; cotton,patient,touching; forceps,cotton,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1045_0", + "timepoint": 1045, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1045, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1803_0", + "timepoint": 1803, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1803, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,needle,holding; assistant,syringe,holding; patient,operating table,lyingon; assistant,instrument table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_302_0", + "timepoint": 302, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 302, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; assistant,operating table,closeto; forceps,patient,touching; head surgeon,curette,holding; patient,operating table,lyingon; head surgeon,forceps,holding; curette,patient,touching; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1221_0", + "timepoint": 1221, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1221, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,ultrasound machine,closeto; circulator,ultrasound machine,manipulating; assistant,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_36_0", + "timepoint": 36, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 36, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_3097_0", + "timepoint": 3097, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3097, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,tissue paper,holding; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_340_0", + "timepoint": 340, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 340, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,forceps,holding; patient,operating table,lyingon; head surgeon,curette,holding; curette,patient,touching; head surgeon,herbal disk,removing; assistant,operating table,closeto; forceps,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2787_0", + "timepoint": 2787, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2787, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_529_0", + "timepoint": 529, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 529, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_746_0", + "timepoint": 746, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 746, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_866_0", + "timepoint": 866, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 866, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_804_0", + "timepoint": 804, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 804, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,forceps,holding; forceps,cotton,touching; patient,operating table,lyingon; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1535_0", + "timepoint": 1535, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1535, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_2273_0", + "timepoint": 2273, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2273, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,forceps,holding; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1047_0", + "timepoint": 1047, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1047, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_388_0", + "timepoint": 388, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 388, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_2_2_776_0", + "timepoint": 776, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 776, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_298_0", + "timepoint": 298, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 298, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1431_0", + "timepoint": 1431, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1431, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,removing; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_732_0", + "timepoint": 732, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 732, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_813_0", + "timepoint": 813, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 813, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2894_0", + "timepoint": 2894, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2894, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,syringe,removing; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_789_0", + "timepoint": 789, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 789, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_3_6_76_0", + "timepoint": 76, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 76, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,curette,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2253_0", + "timepoint": 2253, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2253, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; syringe,needle,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,needle,holding; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_2_2_549_0", + "timepoint": 549, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 549, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; anesthetist,operating table,closeto; anesthetist,health monitor,looking; head surgeon,microscope controller,controlling; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2074_0", + "timepoint": 2074, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2074, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; syringe,needle,touching; needle,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1506_0", + "timepoint": 1506, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1506, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_460_0", + "timepoint": 460, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 460, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_100_0", + "timepoint": 100, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 100, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,curette,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2466_0", + "timepoint": 2466, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2466, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; assistant,patient,manipulating; assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_581_0", + "timepoint": 581, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 581, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_671_0", + "timepoint": 671, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 671, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_680_0", + "timepoint": 680, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 680, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_213_0", + "timepoint": 213, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 213, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_382_0", + "timepoint": 382, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 382, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; head surgeon,operating table,closeto; assistant,instrument table,looking; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_2_2_535_0", + "timepoint": 535, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 535, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,controlling; anesthetist,health monitor,looking; circulator,instrument table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_265_0", + "timepoint": 265, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 265, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,ultrasound gel,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_849_0", + "timepoint": 849, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 849, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; assistant,forceps,holding; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,antiseptic,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1171_0", + "timepoint": 1171, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1171, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_2318_0", + "timepoint": 2318, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2318, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2729_0", + "timepoint": 2729, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2729, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,instrument table,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding; assistant,operating table,closeto; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2299_0", + "timepoint": 2299, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2299, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,syringe,holding; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1589_0", + "timepoint": 1589, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1589, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_539_0", + "timepoint": 539, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 539, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_2_2_1318_0", + "timepoint": 1318, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1318, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,antiseptic,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2355_0", + "timepoint": 2355, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2355, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,syringe,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1282_0", + "timepoint": 1282, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1282, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,applying; body marker,patient,touching; ultrasound probe,patient,touching; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2266_0", + "timepoint": 2266, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2266, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound probe,controlling; assistant,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2584_0", + "timepoint": 2584, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2584, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; assistant,needle,holding; head surgeon,vertebrae,scanning; syringe,needle,touching; needle,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_812_0", + "timepoint": 812, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 812, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; assistant,operating table,closeto; forceps,cotton,touching; assistant,antiseptic,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1824_0", + "timepoint": 1824, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1824, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,antiseptic,holding; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1191_0", + "timepoint": 1191, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1191, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,ultrasound probe,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_471_0", + "timepoint": 471, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 471, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2456_0", + "timepoint": 2456, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2456, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,instrument table,closeto; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1486_0", + "timepoint": 1486, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1486, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1813_0", + "timepoint": 1813, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1813, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3166_0", + "timepoint": 3166, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3166, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; assistant,tissue paper,applying; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1489_0", + "timepoint": 1489, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1489, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2392_0", + "timepoint": 2392, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2392, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,syringe,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1473_0", + "timepoint": 1473, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1473, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,patient,anaesthetising; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope controller,manipulating; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2949_0", + "timepoint": 2949, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2949, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,scanning; needle,patient,touching; assistant,syringe,holding " + } + ] + }, + { + "id": "MISS_1_1_197_0", + "timepoint": 197, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 197, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_874_0", + "timepoint": 874, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 874, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1883_0", + "timepoint": 1883, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1883, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,patient,disinfection; assistant,antiseptic,applying " + } + ] + }, + { + "id": "MISS_1_1_1358_0", + "timepoint": 1358, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1358, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,operating table,closeto; assistant,instrument table,looking; anesthetist,anesthesia equipment,controlling; anesthetist,patient,anaesthetising; circulator,operating table,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1934_0", + "timepoint": 1934, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1934, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; assistant,antiseptic,applying; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2496_0", + "timepoint": 2496, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2496, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2641_0", + "timepoint": 2641, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2641, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_902_0", + "timepoint": 902, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 902, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,forceps,holding; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2277_0", + "timepoint": 2277, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2277, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2744_0", + "timepoint": 2744, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2744, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; syringe,needle,touching; assistant,operating table,closeto; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1475_0", + "timepoint": 1475, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1475, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,patient,anaesthetising; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope controller,manipulating; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_673_0", + "timepoint": 673, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 673, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,curette,holding; assistant,instrument table,closeto; assistant,instrument table,looking; assistant,scalpel,holding; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_108_0", + "timepoint": 108, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 108, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1183_0", + "timepoint": 1183, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1183, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating room,entering; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_283_0", + "timepoint": 283, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 283, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,manipulating; patient,operating room,entering; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_737_0", + "timepoint": 737, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 737, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1789_0", + "timepoint": 1789, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1789, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_867_0", + "timepoint": 867, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 867, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; body marker,patient,touching; assistant,operating table,closeto; assistant,body marker,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_487_0", + "timepoint": 487, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 487, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_830_0", + "timepoint": 830, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 830, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1911_0", + "timepoint": 1911, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1911, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,needle,applying " + } + ] + }, + { + "id": "Ultrasound_2_2_1319_0", + "timepoint": 1319, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1319, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,antiseptic,holding; patient,operating table,lyingon; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_283_0", + "timepoint": 283, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 283, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2132_0", + "timepoint": 2132, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2132, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,needle,holding; syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_3_4_581_0", + "timepoint": 581, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 581, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2236_0", + "timepoint": 2236, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2236, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; cotton,patient,touching; head surgeon,operating table,closeto; assistant,patient,disinfection; forceps,cotton,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_967_0", + "timepoint": 967, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 967, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; patient,operating table,lyingon; assistant,antiseptic,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,disinfection; forceps,cotton,touching; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_206_0", + "timepoint": 206, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 206, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1448_0", + "timepoint": 1448, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1448, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2839_0", + "timepoint": 2839, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2839, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; assistant,syringe,removing; syringe,needle,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_876_0", + "timepoint": 876, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 876, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,microscope,manipulating; anesthetist,operating table,closeto; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1933_0", + "timepoint": 1933, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1933, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,ultrasound probe,looking; head surgeon,ultrasound screen,looking; assistant,needle,inserting; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,syringe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1541_0", + "timepoint": 1541, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1541, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_186_0", + "timepoint": 186, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 186, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_47_0", + "timepoint": 47, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 47, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; assistant,instrument table,closeto; assistant,instrument table,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_556_0", + "timepoint": 556, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 556, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating room,entering; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_945_0", + "timepoint": 945, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 945, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1896_0", + "timepoint": 1896, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1896, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1599_0", + "timepoint": 1599, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1599, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; assistant,needle,inserting; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_91_0", + "timepoint": 91, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 91, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling; circulator,body marker,looking " + } + ] + }, + { + "id": "MISS_1_1_1143_0", + "timepoint": 1143, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1143, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating room,entering; assistant,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_193_0", + "timepoint": 193, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 193, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,curette,holding; head surgeon,herbal disk,removing; assistant,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,forceps,holding; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2934_0", + "timepoint": 2934, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2934, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,needle,removing; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2351_0", + "timepoint": 2351, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2351, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,needle,holding; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; syringe,needle,touching; assistant,syringe,injecting; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_279_0", + "timepoint": 279, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 279, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_431_0", + "timepoint": 431, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 431, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "MISS_2_2_251_0", + "timepoint": 251, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 251, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; assistant,forceps,holding; head surgeon,patient,cutting; head surgeon,scalpel,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; scalpel,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1574_0", + "timepoint": 1574, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1574, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_99_0", + "timepoint": 99, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 99, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,curette,holding; curette,patient,touching; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1668_0", + "timepoint": 1668, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1668, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1281_0", + "timepoint": 1281, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1281, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,needle,inserting; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_309_0", + "timepoint": 309, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 309, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_791_0", + "timepoint": 791, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 791, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_224_0", + "timepoint": 224, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 224, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_4_0", + "timepoint": 4, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 4, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_2194_0", + "timepoint": 2194, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2194, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; assistant,operating table,closeto; assistant,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,syringe,injecting; syringe,needle,touching; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1691_0", + "timepoint": 1691, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1691, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,inserting; assistant,tissue mark,looking; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1312_0", + "timepoint": 1312, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1312, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,inserting; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_506_0", + "timepoint": 506, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 506, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_2367_0", + "timepoint": 2367, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2367, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_865_0", + "timepoint": 865, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 865, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1072_0", + "timepoint": 1072, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1072, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,body marker,holding; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_261_0", + "timepoint": 261, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 261, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1777_0", + "timepoint": 1777, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1777, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,instrument table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2877_0", + "timepoint": 2877, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2877, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,syringe,removing; syringe,needle,touching; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; needle,patient,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_288_0", + "timepoint": 288, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 288, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound gel,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1507_0", + "timepoint": 1507, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1507, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_189_0", + "timepoint": 189, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 189, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,curette,holding; curette,patient,touching; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_525_0", + "timepoint": 525, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 525, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_420_0", + "timepoint": 420, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 420, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,touching; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3119_0", + "timepoint": 3119, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3119, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_605_0", + "timepoint": 605, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 605, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_240_0", + "timepoint": 240, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 240, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound gel,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_358_0", + "timepoint": 358, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 358, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound gel,applying; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_256_0", + "timepoint": 256, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 256, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,patient,touching; anesthetist,health monitor,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; curette,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,forceps,holding; anesthetist,operating table,closeto; head surgeon,curette,holding; head surgeon,herbal disk,removing; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1666_0", + "timepoint": 1666, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1666, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue mark,looking; assistant,needle,inserting; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_985_0", + "timepoint": 985, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 985, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_142_0", + "timepoint": 142, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 142, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,scissors,looking; circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_342_0", + "timepoint": 342, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 342, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,looking; patient,operating table,closeto; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_274_0", + "timepoint": 274, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 274, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,patient,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_3126_0", + "timepoint": 3126, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3126, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; tissue paper,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_1198_0", + "timepoint": 1198, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1198, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,needle,inserting; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,ultrasound probe,looking; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_59_0", + "timepoint": 59, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 59, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2019_0", + "timepoint": 2019, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2019, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; syringe,needle,touching; assistant,syringe,injecting; head surgeon,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_727_0", + "timepoint": 727, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 727, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; head surgeon,microscope eye,looking; assistant,antiseptic,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_75_0", + "timepoint": 75, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 75, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1170_0", + "timepoint": 1170, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1170, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2664_0", + "timepoint": 2664, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2664, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; head surgeon,needle,holding; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2269_0", + "timepoint": 2269, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2269, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1143_0", + "timepoint": 1143, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1143, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; circulator,ultrasound machine,closeto; circulator,ultrasound machine,manipulating; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2792_0", + "timepoint": 2792, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2792, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,needle,holding; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_2578_0", + "timepoint": 2578, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2578, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,operating table,closeto; assistant,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_61_0", + "timepoint": 61, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 61, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_418_0", + "timepoint": 418, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 418, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2427_0", + "timepoint": 2427, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2427, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,needle,holding; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_236_0", + "timepoint": 236, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 236, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_640_0", + "timepoint": 640, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 640, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,patient,manipulating; assistant,patient,manipulating; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1699_0", + "timepoint": 1699, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1699, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,inserting; needle,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1405_0", + "timepoint": 1405, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1405, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; assistant,needle,removing; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_578_0", + "timepoint": 578, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 578, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_992_0", + "timepoint": 992, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 992, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,microscope,manipulating; anesthetist,operating table,closeto; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_2122_0", + "timepoint": 2122, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2122, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; syringe,needle,touching; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,needle,holding; patient,operating table,lyingon; assistant,needle,holding; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2377_0", + "timepoint": 2377, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2377, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,syringe,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_2_2_244_0", + "timepoint": 244, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 244, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,scalpel,holding; assistant,forceps,holding; anesthetist,operating table,closeto; scalpel,patient,touching; head surgeon,patient,cutting; anesthetist,anesthesia equipment,controlling; forceps,cotton,touching; patient,operating table,lyingon; anesthetist,health monitor,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_320_0", + "timepoint": 320, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 320, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; curette,patient,touching; head surgeon,forceps,holding; forceps,patient,touching; head surgeon,curette,holding; patient,operating table,lyingon; head surgeon,herbal disk,removing; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "MISS_1_1_502_0", + "timepoint": 502, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 502, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; anesthetist,operating table,manipulating; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_1034_0", + "timepoint": 1034, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1034, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,patient,disinfection; head surgeon,operating table,closeto; cotton,patient,touching; forceps,cotton,touching; assistant,antiseptic,holding; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1117_0", + "timepoint": 1117, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1117, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_361_0", + "timepoint": 361, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 361, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_432_0", + "timepoint": 432, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 432, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; patient,operating table,touching; circulator,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2565_0", + "timepoint": 2565, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2565, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; head surgeon,patient,manipulating; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2665_0", + "timepoint": 2665, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2665, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2344_0", + "timepoint": 2344, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2344, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,removing; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_483_0", + "timepoint": 483, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 483, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; head surgeon,herbal disk,removing; head surgeon,curette,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; scalpel,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_901_0", + "timepoint": 901, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 901, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; circulator,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_723_0", + "timepoint": 723, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 723, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1405_0", + "timepoint": 1405, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1405, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,body marker,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,looking; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3073_0", + "timepoint": 3073, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3073, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,tissue paper,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1806_0", + "timepoint": 1806, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1806, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1395_0", + "timepoint": 1395, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1395, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_179_0", + "timepoint": 179, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 179, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_384_0", + "timepoint": 384, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 384, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,aspirating; cotton,patient,touching; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_982_0", + "timepoint": 982, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 982, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,antiseptic,holding; cotton,patient,touching; assistant,patient,disinfection; forceps,cotton,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_132_0", + "timepoint": 132, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 132, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_881_0", + "timepoint": 881, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 881, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,applying; head surgeon,operating table,closeto; assistant,patient,disinfection; assistant,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2391_0", + "timepoint": 2391, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2391, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,needle,removing; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_5_0", + "timepoint": 5, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 5, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1526_0", + "timepoint": 1526, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1526, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,microscope controller,manipulating; anesthetist,patient,anaesthetising; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_440_0", + "timepoint": 440, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 440, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3071_0", + "timepoint": 3071, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3071, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,holding " + } + ] + }, + { + "id": "MISS_2_2_37_0", + "timepoint": 37, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 37, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,looking; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_482_0", + "timepoint": 482, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 482, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_189_0", + "timepoint": 189, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 189, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1036_0", + "timepoint": 1036, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1036, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_758_0", + "timepoint": 758, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 758, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1930_0", + "timepoint": 1930, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1930, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,antiseptic,applying; assistant,patient,disinfection; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1910_0", + "timepoint": 1910, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1910, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,disinfection; patient,operating table,lyingon; assistant,antiseptic,applying; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2024_0", + "timepoint": 2024, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2024, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,patient,disinfection; head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3174_0", + "timepoint": 3174, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3174, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; tissue paper,patient,touching; assistant,tissue paper,applying " + } + ] + }, + { + "id": "MISS_2_2_633_0", + "timepoint": 633, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 633, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope screen,looking; head surgeon,microscope controller,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,microscope screen,looking; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_574_0", + "timepoint": 574, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 574, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1927_0", + "timepoint": 1927, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1927, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,syringe,injecting; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_856_0", + "timepoint": 856, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 856, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; head surgeon,operating table,closeto; assistant,body marker,applying; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; body marker,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2289_0", + "timepoint": 2289, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2289, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2438_0", + "timepoint": 2438, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2438, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; needle,patient,touching; assistant,operating table,closeto; assistant,needle,holding; head surgeon,needle,holding; syringe,needle,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_2_2_1011_0", + "timepoint": 1011, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1011, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; assistant,operating table,closeto; cotton,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,antiseptic,holding; forceps,cotton,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2640_0", + "timepoint": 2640, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2640, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; syringe,needle,touching; needle,patient,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_567_0", + "timepoint": 567, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 567, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,operating table,closeto; head surgeon,curette,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; curette,patient,touching; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; head surgeon,herbal disk,removing; scalpel,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_73_0", + "timepoint": 73, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 73, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,antiseptic,looking; circulator,instrument table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2136_0", + "timepoint": 2136, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2136, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; syringe,needle,touching; patient,operating table,lyingon; assistant,syringe,injecting; needle,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1107_0", + "timepoint": 1107, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1107, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_609_0", + "timepoint": 609, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 609, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_617_0", + "timepoint": 617, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 617, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_87_0", + "timepoint": 87, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 87, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,curette,looking; circulator,instrument table,closeto; circulator,instruments,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_885_0", + "timepoint": 885, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 885, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; body marker,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,ultrasound probe,looking; assistant,body marker,applying " + } + ] + }, + { + "id": "MISS_1_1_725_0", + "timepoint": 725, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 725, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1166_0", + "timepoint": 1166, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1166, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,operating table,closeto; assistant,antiseptic,holding; circulator,ultrasound machine,manipulating; assistant,forceps,holding; patient,operating table,lyingon; assistant,operating table,closeto; circulator,ultrasound machine,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_731_0", + "timepoint": 731, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 731, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2943_0", + "timepoint": 2943, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2943, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,removing; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; patient,operating table,lyingon; head surgeon,patient,scanning " + } + ] + }, + { + "id": "MISS_3_6_159_0", + "timepoint": 159, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 159, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; curette,patient,touching; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,herbal disk,removing; assistant,forceps,holding; head surgeon,curette,holding; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_828_0", + "timepoint": 828, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 828, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_346_0", + "timepoint": 346, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 346, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_1_1_1464_0", + "timepoint": 1464, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1464, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope controller,manipulating; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2410_0", + "timepoint": 2410, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2410, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; assistant,instrument table,looking; assistant,syringe,holding; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_741_0", + "timepoint": 741, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 741, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2568_0", + "timepoint": 2568, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2568, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_814_0", + "timepoint": 814, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 814, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_181_0", + "timepoint": 181, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 181, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_250_0", + "timepoint": 250, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 250, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1518_0", + "timepoint": 1518, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1518, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound probe,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1101_0", + "timepoint": 1101, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1101, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1863_0", + "timepoint": 1863, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1863, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,antiseptic,applying; patient,operating table,lyingon; assistant,patient,disinfection; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_139_0", + "timepoint": 139, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 139, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; head surgeon,curette,holding; anesthetist,operating table,closeto; curette,patient,touching; head surgeon,herbal disk,removing; assistant,instrument table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_266_0", + "timepoint": 266, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 266, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; scalpel,patient,touching; head surgeon,patient,cutting; assistant,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching; anesthetist,health monitor,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1093_0", + "timepoint": 1093, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1093, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_4_5_1967_0", + "timepoint": 1967, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1967, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; syringe,needle,touching; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_159_0", + "timepoint": 159, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 159, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; head surgeon,patient,cutting; head surgeon,operating table,closeto; anesthetist,operating table,closeto; scalpel,patient,touching; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,instrument table,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2813_0", + "timepoint": 2813, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2813, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound probe,holding; assistant,needle,removing; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1400_0", + "timepoint": 1400, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1400, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,needle,removing; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_232_0", + "timepoint": 232, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 232, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_859_0", + "timepoint": 859, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 859, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2270_0", + "timepoint": 2270, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2270, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,holding; assistant,operating table,closeto; head surgeon,needle,holding; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_22_0", + "timepoint": 22, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 22, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_2244_0", + "timepoint": 2244, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2244, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; assistant,needle,holding; assistant,syringe,injecting; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_387_0", + "timepoint": 387, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 387, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1543_0", + "timepoint": 1543, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1543, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1918_0", + "timepoint": 1918, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1918, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,syringe,holding; assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,needle,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_1755_0", + "timepoint": 1755, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1755, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,instrument table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding; needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2580_0", + "timepoint": 2580, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2580, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; needle,patient,touching; assistant,needle,holding; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_991_0", + "timepoint": 991, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 991, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,instrument table,looking; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,needle,holding; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1651_0", + "timepoint": 1651, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1651, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,tissue mark,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_538_0", + "timepoint": 538, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 538, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_226_0", + "timepoint": 226, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 226, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_3329_0", + "timepoint": 3329, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3329, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1590_0", + "timepoint": 1590, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1590, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_584_0", + "timepoint": 584, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 584, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1863_0", + "timepoint": 1863, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1863, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1379_0", + "timepoint": 1379, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1379, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,needle,removing " + } + ] + }, + { + "id": "Ultrasound_3_4_1973_0", + "timepoint": 1973, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1973, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,instrument table,closeto; head surgeon,needle,holding; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_255_0", + "timepoint": 255, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 255, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1871_0", + "timepoint": 1871, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1871, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,disinfection; patient,operating table,lyingon; assistant,antiseptic,applying; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1674_0", + "timepoint": 1674, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1674, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2432_0", + "timepoint": 2432, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2432, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,needle,holding; assistant,instrument table,looking; head surgeon,ultrasound screen,looking; assistant,syringe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1424_0", + "timepoint": 1424, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1424, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,body marker,holding; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,instrument table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_755_0", + "timepoint": 755, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 755, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_784_0", + "timepoint": 784, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 784, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1118_0", + "timepoint": 1118, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1118, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_599_0", + "timepoint": 599, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 599, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1858_0", + "timepoint": 1858, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1858, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; needle,patient,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1999_0", + "timepoint": 1999, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1999, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,ultrasound probe,looking; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,syringe,holding; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_493_0", + "timepoint": 493, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 493, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,health monitor,looking; assistant,scalpel,holding; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1363_0", + "timepoint": 1363, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1363, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_935_0", + "timepoint": 935, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 935, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; assistant,patient,disinfection; assistant,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2230_0", + "timepoint": 2230, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2230, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,syringe,injecting; syringe,needle,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_708_0", + "timepoint": 708, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 708, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon; assistant,forceps,holding; circulator,ultrasound machine,manipulating; circulator,ultrasound machine,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_823_0", + "timepoint": 823, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 823, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,body marker,holding " + } + ] + }, + { + "id": "MISS_1_1_1969_0", + "timepoint": 1969, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1969, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,instrument table,looking; assistant,antiseptic,holding; assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_4_5_163_0", + "timepoint": 163, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 163, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_118_0", + "timepoint": 118, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 118, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; curette,patient,touching; anesthetist,operating table,closeto; head surgeon,herbal disk,removing; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_470_0", + "timepoint": 470, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 470, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_185_0", + "timepoint": 185, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 185, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2279_0", + "timepoint": 2279, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2279, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; needle,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2098_0", + "timepoint": 2098, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2098, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,needle,holding; syringe,needle,touching; assistant,syringe,injecting; assistant,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2439_0", + "timepoint": 2439, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2439, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,syringe,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; needle,patient,touching; patient,operating table,lyingon; assistant,instrument table,closeto; assistant,instrument table,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_916_0", + "timepoint": 916, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 916, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,microscope eye,looking; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1148_0", + "timepoint": 1148, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1148, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,body marker,holding; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3013_0", + "timepoint": 3013, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3013, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_1708_0", + "timepoint": 1708, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1708, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_534_0", + "timepoint": 534, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 534, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; curette,patient,touching; head surgeon,herbal disk,removing; patient,operating table,lyingon; head surgeon,scalpel,holding; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; scalpel,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_89_0", + "timepoint": 89, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 89, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound gel,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2074_0", + "timepoint": 2074, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2074, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2918_0", + "timepoint": 2918, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2918, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_39_0", + "timepoint": 39, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 39, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1658_0", + "timepoint": 1658, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1658, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,instrument table,looking; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2137_0", + "timepoint": 2137, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2137, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,forceps,holding; assistant,patient,disinfection; cotton,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_448_0", + "timepoint": 448, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 448, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_854_0", + "timepoint": 854, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 854, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,forceps,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_3_6_275_0", + "timepoint": 275, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 275, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; curette,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,forceps,holding; forceps,patient,touching; anesthetist,health monitor,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_639_0", + "timepoint": 639, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 639, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; head surgeon,operating table,closeto; assistant,patient,manipulating; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_940_0", + "timepoint": 940, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 940, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2092_0", + "timepoint": 2092, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2092, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_666_0", + "timepoint": 666, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 666, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_664_0", + "timepoint": 664, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 664, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_251_0", + "timepoint": 251, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 251, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,ultrasound gel,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3074_0", + "timepoint": 3074, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3074, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,tissue paper,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2793_0", + "timepoint": 2793, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2793, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1981_0", + "timepoint": 1981, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1981, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; patient,operating table,lyingon; assistant,antiseptic,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,patient,disinfection; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2397_0", + "timepoint": 2397, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2397, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_377_0", + "timepoint": 377, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 377, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1994_0", + "timepoint": 1994, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1994, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,inserting; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,holding; assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1439_0", + "timepoint": 1439, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1439, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2269_0", + "timepoint": 2269, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2269, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2519_0", + "timepoint": 2519, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2519, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,needle,holding; head surgeon,needle,holding; syringe,needle,touching; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1256_0", + "timepoint": 1256, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1256, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_189_0", + "timepoint": 189, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 189, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_1_1_2436_0", + "timepoint": 2436, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2436, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,manipulating; head surgeon,patient,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_87_0", + "timepoint": 87, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 87, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; assistant,operating table,closeto; head surgeon,herbal disk,removing; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2382_0", + "timepoint": 2382, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2382, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_543_0", + "timepoint": 543, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 543, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_847_0", + "timepoint": 847, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 847, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,body marker,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_2410_0", + "timepoint": 2410, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2410, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_657_0", + "timepoint": 657, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 657, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; head surgeon,patient,manipulating; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_490_0", + "timepoint": 490, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 490, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_3_6_692_0", + "timepoint": 692, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 692, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,curette,holding; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_638_0", + "timepoint": 638, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 638, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,patient,manipulating; assistant,patient,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1669_0", + "timepoint": 1669, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1669, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,tissue mark,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,needle,inserting; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_891_0", + "timepoint": 891, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 891, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,microscope eye,looking; assistant,forceps,holding; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2719_0", + "timepoint": 2719, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2719, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,syringe,holding; patient,operating table,lyingon; needle,patient,touching; head surgeon,needle,holding; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_378_0", + "timepoint": 378, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 378, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,instrument table,looking " + } + ] + }, + { + "id": "MISS_3_6_640_0", + "timepoint": 640, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 640, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,curette,holding; assistant,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1262_0", + "timepoint": 1262, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1262, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; assistant,forceps,holding; patient,operating table,lyingon; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_617_0", + "timepoint": 617, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 617, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; curette,patient,touching; scalpel,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,curette,holding; head surgeon,scalpel,holding; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_152_0", + "timepoint": 152, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 152, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,antiseptic,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_519_0", + "timepoint": 519, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 519, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2158_0", + "timepoint": 2158, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2158, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound screen,looking; syringe,needle,touching; assistant,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_2132_0", + "timepoint": 2132, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2132, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,patient,disinfection; cotton,patient,touching; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_180_0", + "timepoint": 180, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 180, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " scalpel,patient,touching; forceps,cotton,touching; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,scalpel,holding; assistant,forceps,holding; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,patient,cutting " + } + ] + }, + { + "id": "Ultrasound_2_2_1216_0", + "timepoint": 1216, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1216, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; circulator,ultrasound machine,manipulating; assistant,antiseptic,holding; circulator,ultrasound machine,closeto; assistant,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_703_0", + "timepoint": 703, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 703, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,microscope screen,looking; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2690_0", + "timepoint": 2690, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2690, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,needle,holding; assistant,operating table,closeto; head surgeon,needle,holding; needle,patient,touching; syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_288_0", + "timepoint": 288, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 288, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; curette,patient,touching; patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,forceps,holding; assistant,operating table,closeto; forceps,patient,touching; head surgeon,herbal disk,removing; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_38_0", + "timepoint": 38, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 38, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2553_0", + "timepoint": 2553, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2553, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3180_0", + "timepoint": 3180, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3180, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; tissue paper,patient,touching; assistant,operating table,closeto; assistant,tissue paper,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_426_0", + "timepoint": 426, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 426, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; forceps,cotton,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1063_0", + "timepoint": 1063, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1063, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2666_0", + "timepoint": 2666, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2666, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2375_0", + "timepoint": 2375, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2375, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,needle,holding; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_976_0", + "timepoint": 976, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 976, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; assistant,patient,aspirating; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2_0", + "timepoint": 2, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2600_0", + "timepoint": 2600, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2600, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_541_0", + "timepoint": 541, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 541, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_741_0", + "timepoint": 741, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 741, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1351_0", + "timepoint": 1351, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1351, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2419_0", + "timepoint": 2419, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2419, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,removing; head surgeon,ultrasound machine,looking; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2536_0", + "timepoint": 2536, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2536, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2249_0", + "timepoint": 2249, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2249, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,ultrasound screen,looking; needle,patient,touching; patient,operating table,lyingon; syringe,needle,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2899_0", + "timepoint": 2899, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2899, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; needle,patient,touching; patient,operating table,lyingon; assistant,syringe,removing; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_774_0", + "timepoint": 774, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 774, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; circulator,microscope,closeto; anesthetist,operating table,closeto; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2_0", + "timepoint": 2, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1912_0", + "timepoint": 1912, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1912, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; assistant,operating table,closeto; needle,patient,touching; head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1077_0", + "timepoint": 1077, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1077, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,needle,inserting; assistant,ultrasound probe,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1353_0", + "timepoint": 1353, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1353, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,patient,anaesthetising; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1391_0", + "timepoint": 1391, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1391, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,removing; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1490_0", + "timepoint": 1490, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1490, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_685_0", + "timepoint": 685, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 685, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_147_0", + "timepoint": 147, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 147, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,scalpel,holding; scalpel,patient,touching; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,patient,cutting " + } + ] + }, + { + "id": "Ultrasound_5_6_1636_0", + "timepoint": 1636, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1636, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,instrument table,looking; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_32_0", + "timepoint": 32, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 32, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,patient,manipulating; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_582_0", + "timepoint": 582, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 582, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,scalpel,holding; head surgeon,curette,holding; scalpel,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1760_0", + "timepoint": 1760, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1760, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,syringe,holding; head surgeon,ultrasound screen,looking; assistant,instrument table,closeto; assistant,needle,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1027_0", + "timepoint": 1027, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1027, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_197_0", + "timepoint": 197, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 197, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_2_2_483_0", + "timepoint": 483, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 483, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,scalpel,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_296_0", + "timepoint": 296, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 296, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_99_0", + "timepoint": 99, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 99, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1750_0", + "timepoint": 1750, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1750, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,syringe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_711_0", + "timepoint": 711, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 711, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_330_0", + "timepoint": 330, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 330, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering; circulator,operating table,manipulating " + } + ] + }, + { + "id": "MISS_2_2_571_0", + "timepoint": 571, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 571, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,microscope controller,controlling; patient,operating table,lyingon; anesthetist,health monitor,looking; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_790_0", + "timepoint": 790, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 790, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1969_0", + "timepoint": 1969, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1969, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; assistant,syringe,holding; assistant,instrument table,closeto; ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2209_0", + "timepoint": 2209, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2209, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1252_0", + "timepoint": 1252, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1252, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound probe,looking; assistant,needle,inserting; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1155_0", + "timepoint": 1155, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1155, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating room,entering; head surgeon,operating room,entering; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1628_0", + "timepoint": 1628, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1628, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,inserting; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,tissue mark,looking; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_722_0", + "timepoint": 722, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 722, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_659_0", + "timepoint": 659, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 659, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2317_0", + "timepoint": 2317, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2317, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3147_0", + "timepoint": 3147, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3147, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,tissue paper,applying; head surgeon,operating table,closeto; tissue paper,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_558_0", + "timepoint": 558, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 558, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2049_0", + "timepoint": 2049, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2049, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,syringe,injecting; syringe,needle,touching; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_702_0", + "timepoint": 702, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 702, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,curette,holding; assistant,instrument table,closeto; head surgeon,microscope eye,looking; assistant,instrument table,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1111_0", + "timepoint": 1111, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1111, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1639_0", + "timepoint": 1639, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1639, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_904_0", + "timepoint": 904, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 904, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding; assistant,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_2_2_1254_0", + "timepoint": 1254, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1254, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,operating table,closeto; assistant,antiseptic,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_372_0", + "timepoint": 372, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 372, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; head surgeon,forceps,holding; patient,operating table,lyingon; head surgeon,curette,holding; curette,patient,touching; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; forceps,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3296_0", + "timepoint": 3296, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3296, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,bin,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2013_0", + "timepoint": 2013, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2013, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1464_0", + "timepoint": 1464, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1464, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_920_0", + "timepoint": 920, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 920, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_370_0", + "timepoint": 370, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 370, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound gel,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1746_0", + "timepoint": 1746, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1746, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; assistant,syringe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_2_2_443_0", + "timepoint": 443, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 443, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; assistant,cotton,holding; anesthetist,operating table,closeto; assistant,instrument table,looking; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2430_0", + "timepoint": 2430, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2430, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,vertebrae,scanning; assistant,syringe,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,instrument table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_1281_0", + "timepoint": 1281, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1281, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,operating table,closeto; assistant,antiseptic,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_2239_0", + "timepoint": 2239, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2239, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; assistant,forceps,holding; assistant,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_3_4_2176_0", + "timepoint": 2176, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2176, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; needle,patient,touching; syringe,needle,touching; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_3_4_2306_0", + "timepoint": 2306, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2306, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,needle,holding; assistant,syringe,injecting; assistant,needle,holding; head surgeon,ultrasound screen,looking; syringe,needle,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2469_0", + "timepoint": 2469, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2469, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,patient,manipulating; anesthetist,operating table,closeto; assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1942_0", + "timepoint": 1942, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1942, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; needle,patient,touching; assistant,syringe,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1984_0", + "timepoint": 1984, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1984, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; needle,patient,touching; patient,operating table,lyingon; assistant,syringe,holding; head surgeon,needle,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_669_0", + "timepoint": 669, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 669, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1910_0", + "timepoint": 1910, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1910, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,syringe,holding; assistant,needle,applying; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2041_0", + "timepoint": 2041, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2041, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,needle,inserting; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,syringe,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_147_0", + "timepoint": 147, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 147, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound machine,looking; assistant,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_80_0", + "timepoint": 80, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 80, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_493_0", + "timepoint": 493, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 493, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1309_0", + "timepoint": 1309, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1309, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_219_0", + "timepoint": 219, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 219, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,scalpel,holding; head surgeon,patient,cutting; patient,operating table,lyingon; head surgeon,operating table,closeto; scalpel,patient,touching; assistant,operating table,closeto; anesthetist,health monitor,looking; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_73_0", + "timepoint": 73, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 73, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound machine,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_71_0", + "timepoint": 71, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 71, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_3_6_589_0", + "timepoint": 589, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 589, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; curette,patient,touching; head surgeon,curette,holding; anesthetist,operating table,closeto; head surgeon,herbal disk,removing; scalpel,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2490_0", + "timepoint": 2490, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2490, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,needle,holding; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2040_0", + "timepoint": 2040, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2040, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; assistant,instrument table,closeto; assistant,patient,disinfection; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_124_0", + "timepoint": 124, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 124, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,scalpel,looking; circulator,instrument table,closeto; circulator,instruments,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1811_0", + "timepoint": 1811, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1811, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; assistant,syringe,holding; head surgeon,needle,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1936_0", + "timepoint": 1936, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1936, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,injecting; needle,patient,touching; syringe,needle,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_2320_0", + "timepoint": 2320, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2320, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1794_0", + "timepoint": 1794, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1794, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,inserting; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,tissue mark,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1302_0", + "timepoint": 1302, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1302, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,patient,manipulating; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_559_0", + "timepoint": 559, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 559, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3132_0", + "timepoint": 3132, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3132, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,applying; tissue paper,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_550_0", + "timepoint": 550, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 550, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_187_0", + "timepoint": 187, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 187, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_662_0", + "timepoint": 662, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 662, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_289_0", + "timepoint": 289, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 289, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; anesthetist,operating table,closeto; curette,patient,touching; head surgeon,forceps,holding; head surgeon,operating table,closeto; head surgeon,curette,holding; forceps,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_48_0", + "timepoint": 48, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 48, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope eye,looking; assistant,curette,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1610_0", + "timepoint": 1610, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1610, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_982_0", + "timepoint": 982, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 982, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; cotton,patient,touching; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,aspirating " + } + ] + }, + { + "id": "MISS_1_1_313_0", + "timepoint": 313, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 313, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_873_0", + "timepoint": 873, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 873, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1877_0", + "timepoint": 1877, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1877, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; needle,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1139_0", + "timepoint": 1139, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1139, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_465_0", + "timepoint": 465, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 465, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_3_6_200_0", + "timepoint": 200, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 200, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; patient,operating table,lyingon; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,herbal disk,removing; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,curette,holding; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_129_0", + "timepoint": 129, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 129, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3159_0", + "timepoint": 3159, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3159, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; tissue paper,patient,touching; assistant,operating table,closeto; assistant,tissue paper,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1041_0", + "timepoint": 1041, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1041, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_2113_0", + "timepoint": 2113, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2113, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,patient,disinfection; assistant,forceps,holding; cotton,patient,touching; forceps,cotton,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1318_0", + "timepoint": 1318, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1318, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_527_0", + "timepoint": 527, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 527, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; head surgeon,curette,holding; anesthetist,operating table,closeto; curette,patient,touching; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; head surgeon,herbal disk,removing; patient,operating table,lyingon; scalpel,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2352_0", + "timepoint": 2352, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2352, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; needle,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,needle,holding; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_494_0", + "timepoint": 494, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 494, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_282_0", + "timepoint": 282, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 282, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_175_0", + "timepoint": 175, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 175, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_1_1_60_0", + "timepoint": 60, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 60, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,curette,looking; circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_480_0", + "timepoint": 480, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 480, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; anesthetist,operating table,manipulating; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_1036_0", + "timepoint": 1036, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1036, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,antiseptic,holding; patient,operating table,lyingon; assistant,operating table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto; cotton,patient,touching; assistant,patient,disinfection " + } + ] + }, + { + "id": "MISS_1_1_907_0", + "timepoint": 907, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 907, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,microscope,manipulating " + } + ] + }, + { + "id": "MISS_3_6_319_0", + "timepoint": 319, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 319, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,microscope eye,looking; forceps,patient,touching; anesthetist,health monitor,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,forceps,holding; head surgeon,herbal disk,removing; patient,operating table,lyingon; curette,patient,touching; head surgeon,curette,holding " + } + ] + }, + { + "id": "MISS_1_1_1945_0", + "timepoint": 1945, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1945, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,applying; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,patient,disinfection; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1087_0", + "timepoint": 1087, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1087, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; anesthetist,operating table,closeto; circulator,microscope,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_671_0", + "timepoint": 671, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 671, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,scalpel,holding; head surgeon,microscope eye,looking; assistant,curette,holding; assistant,instrument table,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1728_0", + "timepoint": 1728, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1728, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2617_0", + "timepoint": 2617, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2617, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_28_0", + "timepoint": 28, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 28, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1415_0", + "timepoint": 1415, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1415, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,patient,anaesthetising; patient,operating table,lyingon; head surgeon,microscope controller,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_21_0", + "timepoint": 21, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 21, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1835_0", + "timepoint": 1835, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1835, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,syringe,injecting; patient,operating table,lyingon; needle,patient,touching; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_324_0", + "timepoint": 324, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 324, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_919_0", + "timepoint": 919, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 919, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,body marker,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1310_0", + "timepoint": 1310, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1310, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound probe,looking; body marker,patient,touching; assistant,body marker,applying; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1399_0", + "timepoint": 1399, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1399, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,patient,anaesthetising; head surgeon,microscope controller,holding; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_2339_0", + "timepoint": 2339, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2339, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_228_0", + "timepoint": 228, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 228, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1876_0", + "timepoint": 1876, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1876, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1493_0", + "timepoint": 1493, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1493, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,inserting; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "MISS_1_1_2193_0", + "timepoint": 2193, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2193, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; forceps,cotton,touching; cotton,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; assistant,patient,disinfection; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2379_0", + "timepoint": 2379, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2379, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,syringe,injecting; ultrasound probe,patient,touching; assistant,needle,holding; needle,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; syringe,needle,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1982_0", + "timepoint": 1982, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1982, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching; assistant,needle,inserting; assistant,syringe,holding; assistant,ultrasound probe,looking; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_555_0", + "timepoint": 555, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 555, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_734_0", + "timepoint": 734, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 734, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_103_0", + "timepoint": 103, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 103, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_499_0", + "timepoint": 499, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 499, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; assistant,scalpel,holding; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_985_0", + "timepoint": 985, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 985, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; forceps,cotton,touching; assistant,patient,disinfection; assistant,forceps,holding; assistant,antiseptic,holding; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1070_0", + "timepoint": 1070, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1070, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,antiseptic,holding; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; circulator,ultrasound machine,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_523_0", + "timepoint": 523, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 523, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2777_0", + "timepoint": 2777, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2777, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,operating table,closeto; assistant,needle,removing; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_373_0", + "timepoint": 373, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 373, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_34_0", + "timepoint": 34, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 34, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound machine,looking; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_455_0", + "timepoint": 455, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 455, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1283_0", + "timepoint": 1283, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1283, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; assistant,needle,inserting " + } + ] + }, + { + "id": "MISS_1_1_777_0", + "timepoint": 777, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 777, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; circulator,microscope,manipulating; circulator,microscope,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_169_0", + "timepoint": 169, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 169, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound machine,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_604_0", + "timepoint": 604, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 604, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_647_0", + "timepoint": 647, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 647, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,controlling; assistant,operating table,closeto; head surgeon,microscope screen,looking; head surgeon,operating table,closeto; assistant,microscope screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1679_0", + "timepoint": 1679, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1679, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,needle,inserting; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,tissue mark,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_31_0", + "timepoint": 31, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 31, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_1_1_1234_0", + "timepoint": 1234, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1234, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,patient,manipulating; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2037_0", + "timepoint": 2037, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2037, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; needle,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2612_0", + "timepoint": 2612, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2612, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1341_0", + "timepoint": 1341, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1341, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; anesthetist,patient,anaesthetising; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_274_0", + "timepoint": 274, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 274, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_780_0", + "timepoint": 780, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 780, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_1876_0", + "timepoint": 1876, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1876, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,antiseptic,applying; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,patient,disinfection; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2368_0", + "timepoint": 2368, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2368, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,needle,removing; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2535_0", + "timepoint": 2535, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2535, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,manipulating " + } + ] + }, + { + "id": "MISS_2_2_755_0", + "timepoint": 755, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 755, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,microscope screen,looking; assistant,microscope screen,looking; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2198_0", + "timepoint": 2198, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2198, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,patient,disinfection; anesthetist,operating table,closeto; forceps,cotton,touching; assistant,operating table,closeto; cotton,patient,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_3_6_986_0", + "timepoint": 986, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 986, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,microscope eye,looking; assistant,patient,aspirating; patient,operating table,lyingon; anesthetist,operating table,closeto; cotton,patient,touching; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_257_0", + "timepoint": 257, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 257, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2114_0", + "timepoint": 2114, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2114, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,vertebrae,scanning; needle,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; syringe,needle,touching; assistant,syringe,injecting; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_242_0", + "timepoint": 242, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 242, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2669_0", + "timepoint": 2669, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2669, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,needle,holding; patient,operating table,lyingon; assistant,syringe,injecting; syringe,needle,touching; assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_782_0", + "timepoint": 782, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 782, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_848_0", + "timepoint": 848, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 848, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_718_0", + "timepoint": 718, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 718, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_710_0", + "timepoint": 710, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 710, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; circulator,ultrasound machine,manipulating; circulator,ultrasound machine,closeto; assistant,forceps,holding; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "MISS_1_1_1344_0", + "timepoint": 1344, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1344, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,patient,anaesthetising " + } + ] + }, + { + "id": "Ultrasound_4_5_1300_0", + "timepoint": 1300, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1300, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1013_0", + "timepoint": 1013, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1013, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_985_0", + "timepoint": 985, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 985, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,body marker,holding " + } + ] + }, + { + "id": "MISS_3_6_655_0", + "timepoint": 655, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 655, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,scalpel,holding; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,curette,holding; head surgeon,microscope eye,looking; assistant,instrument table,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_538_0", + "timepoint": 538, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 538, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_539_0", + "timepoint": 539, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 539, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; head surgeon,operating table,closeto; scalpel,patient,touching; curette,patient,touching; assistant,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; head surgeon,scalpel,holding; anesthetist,operating table,closeto; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "MISS_2_2_229_0", + "timepoint": 229, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 229, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,operating table,closeto; scalpel,patient,touching; head surgeon,operating table,closeto; head surgeon,patient,cutting; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,health monitor,looking; anesthetist,operating table,closeto; head surgeon,scalpel,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_681_0", + "timepoint": 681, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 681, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2882_0", + "timepoint": 2882, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2882, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound machine,looking; assistant,needle,holding; assistant,instrument table,looking " + } + ] + }, + { + "id": "MISS_3_6_772_0", + "timepoint": 772, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 772, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2564_0", + "timepoint": 2564, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2564, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_766_0", + "timepoint": 766, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 766, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2148_0", + "timepoint": 2148, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2148, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; forceps,cotton,touching; head surgeon,operating table,closeto; cotton,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; assistant,forceps,holding; assistant,patient,disinfection " + } + ] + }, + { + "id": "MISS_1_1_1032_0", + "timepoint": 1032, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1032, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; patient,operating table,lyingon; circulator,microscope,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1438_0", + "timepoint": 1438, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1438, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_361_0", + "timepoint": 361, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 361, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_2876_0", + "timepoint": 2876, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2876, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,needle,holding; assistant,instrument table,looking; head surgeon,ultrasound machine,looking; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_971_0", + "timepoint": 971, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 971, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1271_0", + "timepoint": 1271, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1271, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_323_0", + "timepoint": 323, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 323, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound gel,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1561_0", + "timepoint": 1561, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1561, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2216_0", + "timepoint": 2216, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2216, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; syringe,needle,touching; patient,operating table,lyingon; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_63_0", + "timepoint": 63, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 63, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_2827_0", + "timepoint": 2827, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2827, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound machine,looking; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,removing; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1926_0", + "timepoint": 1926, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1926, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,patient,disinfection; patient,operating table,lyingon; assistant,operating table,closeto; assistant,antiseptic,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_2556_0", + "timepoint": 2556, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2556, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,tissue paper,applying; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_189_0", + "timepoint": 189, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 189, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2021_0", + "timepoint": 2021, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2021, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,holding; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1118_0", + "timepoint": 1118, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1118, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,closeto; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,operating table,closeto; circulator,ultrasound machine,manipulating; patient,operating table,lyingon; assistant,forceps,holding; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2511_0", + "timepoint": 2511, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2511, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,syringe,injecting; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_730_0", + "timepoint": 730, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 730, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "MISS_3_6_793_0", + "timepoint": 793, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 793, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1000_0", + "timepoint": 1000, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1000, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2069_0", + "timepoint": 2069, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2069, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_970_0", + "timepoint": 970, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 970, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,body marker,holding; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1177_0", + "timepoint": 1177, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1177, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,body marker,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_340_0", + "timepoint": 340, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 340, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,looking; patient,operating table,closeto; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2599_0", + "timepoint": 2599, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2599, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2514_0", + "timepoint": 2514, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2514, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon; assistant,operating table,closeto; assistant,patient,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_990_0", + "timepoint": 990, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 990, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,needle,holding; head surgeon,ultrasound screen,looking; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1113_0", + "timepoint": 1113, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1113, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,forceps,holding; assistant,instrument table,closeto; assistant,cotton,holding; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_3261_0", + "timepoint": 3261, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3261, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,applying; assistant,operating table,closeto; tissue paper,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_344_0", + "timepoint": 344, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 344, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; anesthetist,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_1055_0", + "timepoint": 1055, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1055, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,body marker,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_9_0", + "timepoint": 9, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 9, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_2727_0", + "timepoint": 2727, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2727, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,bin,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_4_0", + "timepoint": 4, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 4, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1364_0", + "timepoint": 1364, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1364, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; circulator,operating table,looking; anesthetist,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_201_0", + "timepoint": 201, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 201, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling " + } + ] + }, + { + "id": "MISS_3_6_306_0", + "timepoint": 306, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 306, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; curette,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; forceps,patient,touching; head surgeon,curette,holding; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; head surgeon,forceps,holding; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "MISS_2_2_417_0", + "timepoint": 417, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 417, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,forceps,holding; forceps,cotton,touching; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_496_0", + "timepoint": 496, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 496, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2211_0", + "timepoint": 2211, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2211, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,needle,holding; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1280_0", + "timepoint": 1280, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1280, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; body marker,patient,touching; assistant,ultrasound probe,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,body marker,applying; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_364_0", + "timepoint": 364, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 364, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; forceps,cotton,touching; assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,scalpel,holding; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; cotton,patient,touching; assistant,patient,aspirating; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1941_0", + "timepoint": 1941, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1941, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,syringe,injecting; needle,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; syringe,needle,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_442_0", + "timepoint": 442, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 442, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1142_0", + "timepoint": 1142, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1142, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,body marker,holding " + } + ] + }, + { + "id": "MISS_1_1_1613_0", + "timepoint": 1613, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1613, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1730_0", + "timepoint": 1730, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1730, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2505_0", + "timepoint": 2505, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2505, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; syringe,needle,touching; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; needle,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_790_0", + "timepoint": 790, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 790, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1289_0", + "timepoint": 1289, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1289, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,inserting; patient,operating table,lyingon; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2104_0", + "timepoint": 2104, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2104, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; patient,operating table,lyingon; assistant,forceps,holding; assistant,operating table,closeto; anesthetist,operating table,closeto; assistant,patient,disinfection; cotton,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_11_0", + "timepoint": 11, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 11, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_73_0", + "timepoint": 73, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 73, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1002_0", + "timepoint": 1002, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1002, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,body marker,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1447_0", + "timepoint": 1447, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1447, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1550_0", + "timepoint": 1550, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1550, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_698_0", + "timepoint": 698, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 698, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,microscope controller,controlling; head surgeon,microscope screen,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; assistant,microscope screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_896_0", + "timepoint": 896, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 896, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,microscope,manipulating; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1991_0", + "timepoint": 1991, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1991, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; assistant,needle,inserting; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,ultrasound probe,looking; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1143_0", + "timepoint": 1143, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1143, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,inserting; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_225_0", + "timepoint": 225, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 225, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling " + } + ] + }, + { + "id": "MISS_2_2_198_0", + "timepoint": 198, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 198, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,health monitor,looking; assistant,operating table,closeto; forceps,cotton,touching; head surgeon,patient,cutting; scalpel,patient,touching; anesthetist,operating table,closeto; assistant,forceps,holding; head surgeon,scalpel,holding; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1081_0", + "timepoint": 1081, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1081, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,body marker,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_2_2_148_0", + "timepoint": 148, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 148, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,scalpel,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,patient,cutting; anesthetist,anesthesia equipment,controlling; scalpel,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_957_0", + "timepoint": 957, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 957, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,body marker,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2793_0", + "timepoint": 2793, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2793, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,needle,holding; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2347_0", + "timepoint": 2347, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2347, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,syringe,holding; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2467_0", + "timepoint": 2467, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2467, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,ultrasound machine,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_744_0", + "timepoint": 744, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 744, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,forceps,holding; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_946_0", + "timepoint": 946, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 946, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,body marker,holding; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_5_0", + "timepoint": 5, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 5, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_289_0", + "timepoint": 289, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 289, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2461_0", + "timepoint": 2461, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2461, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,needle,holding; assistant,instrument table,looking; head surgeon,ultrasound machine,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1665_0", + "timepoint": 1665, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1665, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3169_0", + "timepoint": 3169, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3169, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,applying; tissue paper,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2160_0", + "timepoint": 2160, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2160, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; syringe,needle,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3029_0", + "timepoint": 3029, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3029, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,instrument table,closeto; assistant,tissue paper,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_495_0", + "timepoint": 495, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 495, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating room,entering; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1125_0", + "timepoint": 1125, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1125, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,ultrasound machine,closeto; assistant,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; assistant,antiseptic,holding; head surgeon,operating table,closeto; circulator,ultrasound machine,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2572_0", + "timepoint": 2572, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2572, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,tissue paper,applying; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_711_0", + "timepoint": 711, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 711, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_579_0", + "timepoint": 579, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 579, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_537_0", + "timepoint": 537, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 537, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; curette,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,herbal disk,removing; head surgeon,scalpel,holding; scalpel,patient,touching; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "MISS_3_6_1095_0", + "timepoint": 1095, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1095, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,microscope eye,looking; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2644_0", + "timepoint": 2644, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2644, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; assistant,syringe,injecting; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; syringe,needle,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2548_0", + "timepoint": 2548, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2548, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; syringe,needle,touching; assistant,operating table,closeto; assistant,needle,holding; assistant,syringe,injecting; head surgeon,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1655_0", + "timepoint": 1655, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1655, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_522_0", + "timepoint": 522, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 522, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_213_0", + "timepoint": 213, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 213, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,ultrasound gel,looking; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1595_0", + "timepoint": 1595, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1595, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3077_0", + "timepoint": 3077, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3077, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,holding " + } + ] + }, + { + "id": "MISS_3_6_817_0", + "timepoint": 817, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 817, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,forceps,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2852_0", + "timepoint": 2852, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2852, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; patient,operating table,lyingon; assistant,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2858_0", + "timepoint": 2858, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2858, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; assistant,instrument table,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1553_0", + "timepoint": 1553, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1553, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,patient,anaesthetising; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_573_0", + "timepoint": 573, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 573, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_2_2_470_0", + "timepoint": 470, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 470, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,scalpel,holding " + } + ] + }, + { + "id": "MISS_1_1_2291_0", + "timepoint": 2291, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2291, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1502_0", + "timepoint": 1502, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1502, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2493_0", + "timepoint": 2493, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2493, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_503_0", + "timepoint": 503, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 503, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_986_0", + "timepoint": 986, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 986, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,body marker,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_692_0", + "timepoint": 692, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 692, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1516_0", + "timepoint": 1516, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1516, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_827_0", + "timepoint": 827, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 827, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_3_6_170_0", + "timepoint": 170, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 170, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; curette,patient,touching; anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2734_0", + "timepoint": 2734, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2734, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,bin,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2787_0", + "timepoint": 2787, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2787, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; needle,patient,touching; assistant,needle,removing; head surgeon,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1081_0", + "timepoint": 1081, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1081, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; ultrasound probe,patient,touching; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_5_6_504_0", + "timepoint": 504, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 504, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1021_0", + "timepoint": 1021, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1021, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_2302_0", + "timepoint": 2302, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2302, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,needle,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2321_0", + "timepoint": 2321, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2321, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_1_1_248_0", + "timepoint": 248, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 248, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1275_0", + "timepoint": 1275, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1275, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,body marker,applying; body marker,patient,touching; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1036_0", + "timepoint": 1036, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1036, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,microscope,manipulating; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_841_0", + "timepoint": 841, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 841, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,antiseptic,holding; forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2726_0", + "timepoint": 2726, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2726, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,syringe,holding; needle,patient,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2053_0", + "timepoint": 2053, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2053, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,vertebrae,scanning; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; needle,patient,touching; syringe,needle,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2155_0", + "timepoint": 2155, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2155, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; syringe,needle,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1711_0", + "timepoint": 1711, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1711, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1172_0", + "timepoint": 1172, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1172, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating room,entering; anesthetist,operating table,closeto; head surgeon,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_831_0", + "timepoint": 831, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 831, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_443_0", + "timepoint": 443, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 443, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope eye,looking; assistant,scalpel,holding; head surgeon,forceps,holding; head surgeon,curette,holding; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_3_6_911_0", + "timepoint": 911, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 911, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; forceps,cotton,touching; assistant,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1189_0", + "timepoint": 1189, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1189, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,antiseptic,holding; forceps,cotton,touching; assistant,operating table,closeto; circulator,ultrasound machine,manipulating; head surgeon,operating table,closeto; patient,operating table,lyingon; circulator,ultrasound machine,closeto " + } + ] + }, + { + "id": "MISS_3_6_1083_0", + "timepoint": 1083, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1083, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,forceps,holding; anesthetist,health monitor,looking; forceps,cotton,touching; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_858_0", + "timepoint": 858, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 858, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; assistant,antiseptic,applying; assistant,operating table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_258_0", + "timepoint": 258, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 258, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1404_0", + "timepoint": 1404, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1404, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,body marker,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1048_0", + "timepoint": 1048, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1048, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,body marker,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2598_0", + "timepoint": 2598, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2598, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_628_0", + "timepoint": 628, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 628, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1460_0", + "timepoint": 1460, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1460, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1838_0", + "timepoint": 1838, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1838, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,needle,holding; patient,operating table,lyingon; syringe,needle,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_476_0", + "timepoint": 476, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 476, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,curette,holding; scalpel,patient,touching; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; curette,patient,touching; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,forceps,holding; head surgeon,operating table,closeto; head surgeon,scalpel,holding; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "MISS_1_1_993_0", + "timepoint": 993, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 993, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; anesthetist,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2995_0", + "timepoint": 2995, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2995, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,patient,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,instrument table,looking; patient,operating table,lyingon; assistant,syringe,holding; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2320_0", + "timepoint": 2320, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2320, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2564_0", + "timepoint": 2564, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2564, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; needle,patient,touching; assistant,syringe,injecting; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1763_0", + "timepoint": 1763, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1763, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,tissue mark,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1546_0", + "timepoint": 1546, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1546, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2332_0", + "timepoint": 2332, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2332, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2196_0", + "timepoint": 2196, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2196, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,syringe,injecting; syringe,needle,touching; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2263_0", + "timepoint": 2263, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2263, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; syringe,needle,touching; needle,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_472_0", + "timepoint": 472, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 472, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_321_0", + "timepoint": 321, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 321, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; forceps,cotton,touching; assistant,forceps,holding; assistant,patient,aspirating; patient,operating table,lyingon; head surgeon,operating table,closeto; cotton,patient,touching; anesthetist,health monitor,looking; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3073_0", + "timepoint": 3073, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3073, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_445_0", + "timepoint": 445, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 445, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2392_0", + "timepoint": 2392, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2392, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,needle,removing " + } + ] + }, + { + "id": "Ultrasound_4_5_552_0", + "timepoint": 552, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 552, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1569_0", + "timepoint": 1569, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1569, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; assistant,needle,inserting; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_19_0", + "timepoint": 19, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 19, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_350_0", + "timepoint": 350, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 350, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,ultrasound gel,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_56_0", + "timepoint": 56, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 56, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking; circulator,scalpel,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_654_0", + "timepoint": 654, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 654, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,manipulating; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_608_0", + "timepoint": 608, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 608, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1307_0", + "timepoint": 1307, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1307, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto; anesthetist,patient,manipulating; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2855_0", + "timepoint": 2855, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2855, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1626_0", + "timepoint": 1626, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1626, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2722_0", + "timepoint": 2722, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2722, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,syringe,holding; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,instrument table,looking; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1867_0", + "timepoint": 1867, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1867, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,syringe,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_855_0", + "timepoint": 855, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 855, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; anesthetist,operating table,closeto; anesthetist,microscope,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_147_0", + "timepoint": 147, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 147, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto; circulator,scissors,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1842_0", + "timepoint": 1842, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1842, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2533_0", + "timepoint": 2533, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2533, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,holding; needle,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; syringe,needle,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_154_0", + "timepoint": 154, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 154, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; assistant,forceps,holding; head surgeon,patient,cutting; scalpel,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_76_0", + "timepoint": 76, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 76, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_593_0", + "timepoint": 593, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 593, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1004_0", + "timepoint": 1004, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1004, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1165_0", + "timepoint": 1165, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1165, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,body marker,holding; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2564_0", + "timepoint": 2564, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2564, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,patient,manipulating; assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1241_0", + "timepoint": 1241, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1241, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,inserting; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1435_0", + "timepoint": 1435, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1435, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,needle,removing; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_43_0", + "timepoint": 43, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 43, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; assistant,curette,holding; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_693_0", + "timepoint": 693, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 693, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1161_0", + "timepoint": 1161, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1161, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_710_0", + "timepoint": 710, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 710, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,instrument table,looking; assistant,curette,holding; patient,operating table,lyingon; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_741_0", + "timepoint": 741, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 741, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_995_0", + "timepoint": 995, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 995, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; forceps,cotton,touching; assistant,antiseptic,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_4_5_565_0", + "timepoint": 565, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 565, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1144_0", + "timepoint": 1144, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1144, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,inserting; assistant,ultrasound probe,looking; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1462_0", + "timepoint": 1462, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1462, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_2275_0", + "timepoint": 2275, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2275, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_701_0", + "timepoint": 701, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 701, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1079_0", + "timepoint": 1079, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1079, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2419_0", + "timepoint": 2419, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2419, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,syringe,injecting; needle,patient,touching; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,needle,holding; syringe,needle,touching; head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_337_0", + "timepoint": 337, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 337, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2623_0", + "timepoint": 2623, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2623, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_3213_0", + "timepoint": 3213, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3213, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; tissue paper,patient,touching; assistant,tissue paper,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1460_0", + "timepoint": 1460, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1460, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3106_0", + "timepoint": 3106, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3106, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2053_0", + "timepoint": 2053, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2053, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; needle,patient,touching; syringe,needle,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1018_0", + "timepoint": 1018, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1018, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_270_0", + "timepoint": 270, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 270, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_357_0", + "timepoint": 357, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 357, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,ultrasound gel,applying; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1086_0", + "timepoint": 1086, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1086, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding; assistant,forceps,holding; forceps,cotton,touching; circulator,ultrasound machine,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_113_0", + "timepoint": 113, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 113, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_398_0", + "timepoint": 398, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 398, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,instrument table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "MISS_1_1_586_0", + "timepoint": 586, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 586, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2902_0", + "timepoint": 2902, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2902, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; syringe,needle,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,syringe,removing; head surgeon,ultrasound probe,controlling; needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1821_0", + "timepoint": 1821, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1821, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; needle,patient,touching; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_155_0", + "timepoint": 155, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 155, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound machine,looking; assistant,operating table,closeto; assistant,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_735_0", + "timepoint": 735, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 735, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_824_0", + "timepoint": 824, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 824, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_552_0", + "timepoint": 552, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 552, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1063_0", + "timepoint": 1063, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1063, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,microscope,manipulating; circulator,microscope,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1674_0", + "timepoint": 1674, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1674, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,instrument table,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_943_0", + "timepoint": 943, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 943, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; circulator,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1820_0", + "timepoint": 1820, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1820, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue mark,looking; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_2_2_1342_0", + "timepoint": 1342, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1342, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,forceps,holding; assistant,antiseptic,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_721_0", + "timepoint": 721, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 721, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2513_0", + "timepoint": 2513, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2513, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,needle,holding; needle,patient,touching; syringe,needle,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_3_4_1752_0", + "timepoint": 1752, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1752, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; assistant,tissue mark,looking; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1566_0", + "timepoint": 1566, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1566, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; assistant,operating table,closeto; needle,patient,touching; assistant,needle,inserting; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2604_0", + "timepoint": 2604, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2604, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,injecting; assistant,operating table,closeto; needle,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_778_0", + "timepoint": 778, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 778, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; circulator,microscope,manipulating; anesthetist,operating table,closeto; circulator,microscope,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2287_0", + "timepoint": 2287, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2287, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_50_0", + "timepoint": 50, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 50, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2348_0", + "timepoint": 2348, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2348, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_647_0", + "timepoint": 647, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 647, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_587_0", + "timepoint": 587, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 587, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,scalpel,holding; assistant,operating table,closeto; scalpel,patient,touching; anesthetist,anesthesia equipment,controlling; curette,patient,touching; head surgeon,herbal disk,removing; head surgeon,curette,holding; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_941_0", + "timepoint": 941, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 941, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,microscope,manipulating; circulator,microscope,manipulating " + } + ] + }, + { + "id": "MISS_3_6_64_0", + "timepoint": 64, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 64, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,curette,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_3_6_497_0", + "timepoint": 497, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 497, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " scalpel,patient,touching; patient,operating table,lyingon; head surgeon,curette,holding; head surgeon,scalpel,holding; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; curette,patient,touching; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1071_0", + "timepoint": 1071, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1071, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,antiseptic,holding; forceps,cotton,touching; circulator,ultrasound machine,looking; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2702_0", + "timepoint": 2702, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2702, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; syringe,needle,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1896_0", + "timepoint": 1896, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1896, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,needle,holding; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_394_0", + "timepoint": 394, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 394, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,touching; circulator,patient,closeto; circulator,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_432_0", + "timepoint": 432, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 432, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; head surgeon,scalpel,holding; head surgeon,operating table,closeto; assistant,cotton,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1152_0", + "timepoint": 1152, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1152, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating room,entering; head surgeon,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2577_0", + "timepoint": 2577, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2577, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2163_0", + "timepoint": 2163, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2163, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; syringe,needle,touching; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2743_0", + "timepoint": 2743, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2743, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_927_0", + "timepoint": 927, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 927, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,aspirating; patient,operating table,lyingon; cotton,patient,touching; assistant,operating table,closeto; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; forceps,cotton,touching; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1443_0", + "timepoint": 1443, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1443, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_733_0", + "timepoint": 733, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 733, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1183_0", + "timepoint": 1183, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1183, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,manipulating; assistant,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,antiseptic,holding; forceps,cotton,touching; head surgeon,operating table,closeto; circulator,ultrasound machine,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1005_0", + "timepoint": 1005, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1005, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_641_0", + "timepoint": 641, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 641, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_106_0", + "timepoint": 106, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 106, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1342_0", + "timepoint": 1342, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1342, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,looking; assistant,instrument table,closeto; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_889_0", + "timepoint": 889, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 889, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_12_0", + "timepoint": 12, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 12, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_957_0", + "timepoint": 957, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 957, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_641_0", + "timepoint": 641, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 641, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_447_0", + "timepoint": 447, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 447, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,cotton,holding; patient,operating table,lyingon; head surgeon,scalpel,holding; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; assistant,forceps,holding; assistant,instrument table,looking; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1042_0", + "timepoint": 1042, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1042, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound screen,looking; assistant,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1068_0", + "timepoint": 1068, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1068, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; circulator,ultrasound machine,looking; forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1931_0", + "timepoint": 1931, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1931, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; syringe,needle,touching; head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_477_0", + "timepoint": 477, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 477, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1191_0", + "timepoint": 1191, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1191, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1275_0", + "timepoint": 1275, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1275, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,inserting; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2310_0", + "timepoint": 2310, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2310, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_642_0", + "timepoint": 642, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 642, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_1844_0", + "timepoint": 1844, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1844, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,applying; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,patient,disinfection; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_562_0", + "timepoint": 562, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 562, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_597_0", + "timepoint": 597, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 597, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_2245_0", + "timepoint": 2245, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2245, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; syringe,needle,touching; assistant,syringe,injecting; assistant,operating table,closeto; assistant,ultrasound screen,looking; patient,operating table,lyingon; needle,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2631_0", + "timepoint": 2631, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2631, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_153_0", + "timepoint": 153, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 153, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_695_0", + "timepoint": 695, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 695, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,microscope screen,looking; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope controller,controlling " + } + ] + }, + { + "id": "MISS_1_1_1565_0", + "timepoint": 1565, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1565, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,patient,anaesthetising; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1004_0", + "timepoint": 1004, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1004, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_3_6_475_0", + "timepoint": 475, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 475, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; anesthetist,operating table,closeto; head surgeon,curette,holding; head surgeon,forceps,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "Ultrasound_3_4_1558_0", + "timepoint": 1558, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1558, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_745_0", + "timepoint": 745, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 745, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,instruments,disinfection; anesthetist,health monitor,looking; assistant,anesthetist,applying; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1747_0", + "timepoint": 1747, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1747, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,instrument table,looking; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_376_0", + "timepoint": 376, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 376, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,closeto; circulator,patient,closeto; circulator,operating table,closeto; circulator,operating table,looking " + } + ] + }, + { + "id": "MISS_1_1_874_0", + "timepoint": 874, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 874, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,microscope,manipulating " + } + ] + }, + { + "id": "MISS_3_6_74_0", + "timepoint": 74, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 74, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,curette,holding; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_3_6_621_0", + "timepoint": 621, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 621, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,curette,holding; head surgeon,herbal disk,removing; assistant,operating table,closeto; scalpel,patient,touching; head surgeon,operating table,closeto; curette,patient,touching; anesthetist,health monitor,looking; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_2_2_756_0", + "timepoint": 756, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 756, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,microscope screen,looking; assistant,microscope screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1022_0", + "timepoint": 1022, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1022, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_2386_0", + "timepoint": 2386, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2386, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_592_0", + "timepoint": 592, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 592, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_630_0", + "timepoint": 630, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 630, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_512_0", + "timepoint": 512, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 512, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1359_0", + "timepoint": 1359, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1359, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_2_2_681_0", + "timepoint": 681, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 681, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,microscope controller,controlling; assistant,operating table,closeto; patient,operating table,lyingon; assistant,microscope screen,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2192_0", + "timepoint": 2192, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2192, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; head surgeon,operating table,closeto; assistant,operating table,closeto; syringe,needle,touching; head surgeon,needle,holding; assistant,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_681_0", + "timepoint": 681, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 681, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3089_0", + "timepoint": 3089, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3089, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2108_0", + "timepoint": 2108, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2108, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; forceps,cotton,touching; patient,operating table,lyingon; cotton,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,disinfection; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2084_0", + "timepoint": 2084, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2084, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,syringe,injecting; assistant,operating table,closeto; syringe,needle,touching; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_282_0", + "timepoint": 282, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 282, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,patient,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2061_0", + "timepoint": 2061, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2061, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1450_0", + "timepoint": 1450, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1450, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_48_0", + "timepoint": 48, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 48, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1203_0", + "timepoint": 1203, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1203, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; assistant,operating room,entering; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1820_0", + "timepoint": 1820, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1820, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,needle,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1338_0", + "timepoint": 1338, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1338, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,body marker,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1216_0", + "timepoint": 1216, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1216, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; assistant,ultrasound probe,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,needle,inserting " + } + ] + }, + { + "id": "MISS_1_1_984_0", + "timepoint": 984, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 984, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; patient,operating table,lyingon; circulator,microscope,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_510_0", + "timepoint": 510, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 510, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,operating table,manipulating; anesthetist,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_827_0", + "timepoint": 827, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 827, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; assistant,antiseptic,holding; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_958_0", + "timepoint": 958, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 958, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,forceps,holding; assistant,antiseptic,holding; forceps,cotton,touching; assistant,patient,disinfection; cotton,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1667_0", + "timepoint": 1667, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1667, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,instrument table,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_823_0", + "timepoint": 823, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 823, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2192_0", + "timepoint": 2192, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2192, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; cotton,patient,touching; assistant,patient,disinfection; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_561_0", + "timepoint": 561, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 561, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating room,entering " + } + ] + }, + { + "id": "MISS_1_1_811_0", + "timepoint": 811, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 811, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2675_0", + "timepoint": 2675, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2675, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,tissue paper,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2656_0", + "timepoint": 2656, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2656, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2870_0", + "timepoint": 2870, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2870, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,syringe,removing; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1379_0", + "timepoint": 1379, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1379, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,body marker,holding " + } + ] + }, + { + "id": "MISS_3_6_324_0", + "timepoint": 324, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 324, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,forceps,holding; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; curette,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,curette,holding; forceps,patient,touching; head surgeon,herbal disk,removing; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_570_0", + "timepoint": 570, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 570, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2713_0", + "timepoint": 2713, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2713, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,ultrasound probe,controlling; syringe,needle,touching; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2477_0", + "timepoint": 2477, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2477, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,syringe,injecting; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_183_0", + "timepoint": 183, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 183, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound gel,applying; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_933_0", + "timepoint": 933, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 933, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1448_0", + "timepoint": 1448, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1448, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,patient,anaesthetising; head surgeon,microscope controller,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1523_0", + "timepoint": 1523, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1523, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1196_0", + "timepoint": 1196, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1196, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,inserting; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound probe,looking; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_580_0", + "timepoint": 580, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 580, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1136_0", + "timepoint": 1136, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1136, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_2_2_30_0", + "timepoint": 30, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 30, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1527_0", + "timepoint": 1527, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1527, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1247_0", + "timepoint": 1247, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1247, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; assistant,operating table,closeto; patient,operating table,lyingon; assistant,body marker,applying; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1460_0", + "timepoint": 1460, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1460, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_99_0", + "timepoint": 99, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 99, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_360_0", + "timepoint": 360, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 360, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2760_0", + "timepoint": 2760, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2760, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,needle,holding; assistant,syringe,injecting; needle,patient,touching; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_707_0", + "timepoint": 707, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 707, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,curette,holding; assistant,instrument table,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1681_0", + "timepoint": 1681, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1681, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1556_0", + "timepoint": 1556, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1556, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1459_0", + "timepoint": 1459, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1459, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_3_6_753_0", + "timepoint": 753, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 753, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; assistant,instruments,disinfection; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; assistant,anesthetist,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1017_0", + "timepoint": 1017, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1017, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; forceps,cotton,touching; anesthetist,health monitor,looking; assistant,patient,aspirating; patient,operating table,lyingon; assistant,operating table,closeto; assistant,forceps,holding; cotton,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1795_0", + "timepoint": 1795, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1795, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_371_0", + "timepoint": 371, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 371, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound gel,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_1538_0", + "timepoint": 1538, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1538, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_702_0", + "timepoint": 702, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 702, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,looking; circulator,ultrasound machine,looking; head surgeon,patient,manipulating; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2024_0", + "timepoint": 2024, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2024, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,needle,holding; ultrasound probe,patient,touching; syringe,needle,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2050_0", + "timepoint": 2050, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2050, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding; assistant,syringe,injecting; needle,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2249_0", + "timepoint": 2249, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2249, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,needle,holding; assistant,syringe,injecting; syringe,needle,touching; assistant,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1001_0", + "timepoint": 1001, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1001, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,disinfection; cotton,patient,touching; forceps,cotton,touching; patient,operating table,lyingon; assistant,forceps,holding; assistant,antiseptic,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_367_0", + "timepoint": 367, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 367, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_197_0", + "timepoint": 197, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 197, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; patient,operating table,lyingon; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,patient,cutting; scalpel,patient,touching; forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_770_0", + "timepoint": 770, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 770, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_229_0", + "timepoint": 229, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 229, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2778_0", + "timepoint": 2778, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2778, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_426_0", + "timepoint": 426, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 426, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; curette,patient,touching; assistant,instrument table,looking; anesthetist,anesthesia equipment,controlling; head surgeon,forceps,holding; head surgeon,curette,holding; head surgeon,microscope eye,looking; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_3_6_160_0", + "timepoint": 160, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 160, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,herbal disk,removing; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,curette,holding; curette,patient,touching; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2443_0", + "timepoint": 2443, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2443, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound machine,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_915_0", + "timepoint": 915, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 915, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_298_0", + "timepoint": 298, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 298, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,forceps,holding; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; curette,patient,touching; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; forceps,patient,touching; patient,operating table,lyingon; head surgeon,curette,holding; assistant,operating table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "MISS_3_6_598_0", + "timepoint": 598, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 598, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; head surgeon,herbal disk,removing; scalpel,patient,touching; patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,scalpel,holding; anesthetist,health monitor,looking; head surgeon,operating table,closeto; curette,patient,touching; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2014_0", + "timepoint": 2014, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2014, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; syringe,needle,touching; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2567_0", + "timepoint": 2567, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2567, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; head surgeon,patient,manipulating; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1010_0", + "timepoint": 1010, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1010, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_3_6_102_0", + "timepoint": 102, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 102, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,curette,holding; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1881_0", + "timepoint": 1881, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1881, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,operating table,closeto; patient,operating table,lyingon; assistant,syringe,injecting; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_250_0", + "timepoint": 250, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 250, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1667_0", + "timepoint": 1667, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1667, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,tissue mark,looking; ultrasound probe,patient,touching; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_4_5_612_0", + "timepoint": 612, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 612, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3267_0", + "timepoint": 3267, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3267, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; head surgeon,operating table,closeto; assistant,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_860_0", + "timepoint": 860, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 860, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2995_0", + "timepoint": 2995, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2995, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_3_4_954_0", + "timepoint": 954, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 954, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,body marker,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1847_0", + "timepoint": 1847, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1847, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,syringe,holding; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_16_0", + "timepoint": 16, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 16, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_395_0", + "timepoint": 395, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 395, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,curette,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,forceps,holding; head surgeon,herbal disk,removing; assistant,forceps,holding; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1853_0", + "timepoint": 1853, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1853, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; needle,patient,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2133_0", + "timepoint": 2133, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2133, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; forceps,cotton,touching; cotton,patient,touching; patient,operating table,lyingon; assistant,patient,disinfection; assistant,forceps,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1920_0", + "timepoint": 1920, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1920, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,syringe,injecting; needle,patient,touching; head surgeon,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1618_0", + "timepoint": 1618, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1618, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue mark,looking; assistant,needle,inserting; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_37_0", + "timepoint": 37, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 37, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1100_0", + "timepoint": 1100, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1100, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_746_0", + "timepoint": 746, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 746, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; head surgeon,patient,manipulating; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_923_0", + "timepoint": 923, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 923, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,antiseptic,holding; assistant,patient,disinfection; cotton,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1521_0", + "timepoint": 1521, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1521, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2405_0", + "timepoint": 2405, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2405, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,syringe,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,instrument table,looking; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1376_0", + "timepoint": 1376, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1376, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; assistant,needle,removing; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_302_0", + "timepoint": 302, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 302, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_83_0", + "timepoint": 83, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 83, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,curette,looking; circulator,instrument table,closeto; circulator,instruments,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_464_0", + "timepoint": 464, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 464, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_751_0", + "timepoint": 751, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 751, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_622_0", + "timepoint": 622, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 622, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1914_0", + "timepoint": 1914, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1914, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,applying; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2958_0", + "timepoint": 2958, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2958, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,tissue paper,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; tissue paper,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2498_0", + "timepoint": 2498, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2498, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_2649_0", + "timepoint": 2649, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2649, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,syringe,injecting; syringe,needle,touching; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_2_2_105_0", + "timepoint": 105, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 105, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; assistant,instrument table,looking; head surgeon,scalpel,holding; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_400_0", + "timepoint": 400, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 400, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_581_0", + "timepoint": 581, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 581, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2139_0", + "timepoint": 2139, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2139, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; needle,patient,touching; syringe,needle,touching; ultrasound probe,patient,touching; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1332_0", + "timepoint": 1332, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1332, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,body marker,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2474_0", + "timepoint": 2474, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2474, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,holding; assistant,syringe,injecting; head surgeon,vertebrae,scanning; needle,patient,touching; syringe,needle,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2098_0", + "timepoint": 2098, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2098, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,operating table,closeto; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_3_4_1629_0", + "timepoint": 1629, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1629, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,tissue mark,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,inserting " + } + ] + }, + { + "id": "MISS_3_6_1037_0", + "timepoint": 1037, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1037, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,aspirating; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope eye,looking; cotton,patient,touching; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2711_0", + "timepoint": 2711, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2711, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_245_0", + "timepoint": 245, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 245, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,scalpel,holding; scalpel,patient,touching; assistant,forceps,holding; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; head surgeon,patient,cutting " + } + ] + }, + { + "id": "MISS_1_1_2351_0", + "timepoint": 2351, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2351, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1729_0", + "timepoint": 1729, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1729, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1612_0", + "timepoint": 1612, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1612, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,needle,inserting; patient,operating table,lyingon; assistant,tissue mark,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_2_2_226_0", + "timepoint": 226, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 226, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,health monitor,looking; assistant,operating table,closeto; scalpel,patient,touching; assistant,forceps,holding; head surgeon,scalpel,holding; head surgeon,patient,cutting; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1357_0", + "timepoint": 1357, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1357, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,body marker,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_2541_0", + "timepoint": 2541, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2541, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,manipulating; anesthetist,operating table,closeto; head surgeon,patient,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2014_0", + "timepoint": 2014, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2014, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; assistant,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; assistant,needle,inserting; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_462_0", + "timepoint": 462, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 462, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,microscope eye,looking; head surgeon,scalpel,holding; head surgeon,curette,holding; head surgeon,forceps,holding; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_459_0", + "timepoint": 459, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 459, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1449_0", + "timepoint": 1449, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1449, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_135_0", + "timepoint": 135, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 135, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,scalpel,holding; patient,operating table,lyingon; head surgeon,patient,cutting; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; scalpel,patient,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_1_1_1288_0", + "timepoint": 1288, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1288, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto; anesthetist,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_706_0", + "timepoint": 706, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 706, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2889_0", + "timepoint": 2889, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2889, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound machine,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1173_0", + "timepoint": 1173, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1173, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_977_0", + "timepoint": 977, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 977, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1745_0", + "timepoint": 1745, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1745, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,instrument table,looking " + } + ] + }, + { + "id": "MISS_2_2_378_0", + "timepoint": 378, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 378, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; cotton,patient,touching; assistant,forceps,holding; forceps,cotton,touching; assistant,operating table,closeto; assistant,patient,aspirating " + } + ] + }, + { + "id": "Ultrasound_4_5_1086_0", + "timepoint": 1086, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1086, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; assistant,operating table,closeto; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1354_0", + "timepoint": 1354, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1354, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,removing; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_382_0", + "timepoint": 382, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 382, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_358_0", + "timepoint": 358, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 358, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; patient,operating room,entering; circulator,operating table,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1583_0", + "timepoint": 1583, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1583, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,ultrasound screen,looking; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2783_0", + "timepoint": 2783, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2783, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,needle,removing; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3023_0", + "timepoint": 3023, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3023, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,tissue paper,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; tissue paper,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_527_0", + "timepoint": 527, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 527, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2446_0", + "timepoint": 2446, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2446, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound machine,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_762_0", + "timepoint": 762, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 762, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_356_0", + "timepoint": 356, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 356, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2129_0", + "timepoint": 2129, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2129, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,holding; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,needle,holding; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2374_0", + "timepoint": 2374, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2374, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,removing; assistant,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_2_2_203_0", + "timepoint": 203, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 203, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; anesthetist,health monitor,looking; head surgeon,patient,cutting; patient,operating table,lyingon; scalpel,patient,touching; head surgeon,scalpel,holding; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_204_0", + "timepoint": 204, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 204, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,patient,cutting; assistant,operating table,closeto; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; anesthetist,health monitor,looking; anesthetist,operating table,closeto; scalpel,patient,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1122_0", + "timepoint": 1122, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1122, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,closeto; assistant,operating table,closeto; forceps,cotton,touching; assistant,antiseptic,holding; head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; circulator,ultrasound machine,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_1915_0", + "timepoint": 1915, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1915, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_662_0", + "timepoint": 662, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 662, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_798_0", + "timepoint": 798, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 798, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,antiseptic,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_922_0", + "timepoint": 922, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 922, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,microscope,manipulating; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1302_0", + "timepoint": 1302, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1302, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1814_0", + "timepoint": 1814, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1814, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2059_0", + "timepoint": 2059, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2059, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; assistant,needle,holding; head surgeon,needle,holding; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_2_2_354_0", + "timepoint": 354, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 354, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; circulator,operating table,looking; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_2140_0", + "timepoint": 2140, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2140, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,syringe,injecting; assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; syringe,needle,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_110_0", + "timepoint": 110, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 110, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; assistant,instrument table,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_345_0", + "timepoint": 345, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 345, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2560_0", + "timepoint": 2560, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2560, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_2637_0", + "timepoint": 2637, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2637, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1614_0", + "timepoint": 1614, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1614, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,tissue mark,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,needle,inserting; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1188_0", + "timepoint": 1188, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1188, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2426_0", + "timepoint": 2426, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2426, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_571_0", + "timepoint": 571, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 571, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_675_0", + "timepoint": 675, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 675, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1054_0", + "timepoint": 1054, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1054, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2807_0", + "timepoint": 2807, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2807, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,operating table,closeto; syringe,needle,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1314_0", + "timepoint": 1314, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1314, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_4_5_938_0", + "timepoint": 938, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 938, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_590_0", + "timepoint": 590, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 590, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3065_0", + "timepoint": 3065, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3065, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,tissue paper,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2568_0", + "timepoint": 2568, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2568, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,manipulating; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_449_0", + "timepoint": 449, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 449, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2974_0", + "timepoint": 2974, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2974, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,tissue paper,applying; patient,operating table,lyingon; tissue paper,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1995_0", + "timepoint": 1995, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1995, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,patient,disinfection; assistant,instrument table,looking; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_642_0", + "timepoint": 642, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 642, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1741_0", + "timepoint": 1741, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1741, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,tissue mark,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1698_0", + "timepoint": 1698, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1698, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,instrument table,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1524_0", + "timepoint": 1524, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1524, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1711_0", + "timepoint": 1711, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1711, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; assistant,tissue mark,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2520_0", + "timepoint": 2520, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2520, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_41_0", + "timepoint": 41, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 41, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_433_0", + "timepoint": 433, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 433, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; circulator,patient,closeto; patient,operating table,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_686_0", + "timepoint": 686, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 686, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_1211_0", + "timepoint": 1211, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1211, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; assistant,operating table,closeto; circulator,ultrasound machine,manipulating; forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; circulator,ultrasound machine,closeto " + } + ] + }, + { + "id": "MISS_1_1_1940_0", + "timepoint": 1940, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1940, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,antiseptic,applying; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_4_5_2514_0", + "timepoint": 2514, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2514, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2445_0", + "timepoint": 2445, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2445, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,needle,holding; syringe,needle,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_1_1_553_0", + "timepoint": 553, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 553, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_419_0", + "timepoint": 419, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 419, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound gel,holding; assistant,instrument table,looking; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_2_2_491_0", + "timepoint": 491, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 491, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,scalpel,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,instrument table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "MISS_1_1_1799_0", + "timepoint": 1799, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1799, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1744_0", + "timepoint": 1744, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1744, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,ultrasound screen,looking; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_308_0", + "timepoint": 308, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 308, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,patient,aspirating; head surgeon,scalpel,holding; assistant,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_3_6_1117_0", + "timepoint": 1117, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1117, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; head surgeon,operating table,closeto; anesthetist,health monitor,looking; patient,operating table,lyingon; head surgeon,microscope eye,looking; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_760_0", + "timepoint": 760, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 760, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_778_0", + "timepoint": 778, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 778, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,health monitor,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1482_0", + "timepoint": 1482, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1482, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2674_0", + "timepoint": 2674, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2674, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_587_0", + "timepoint": 587, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 587, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope controller,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1042_0", + "timepoint": 1042, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1042, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; anesthetist,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_3029_0", + "timepoint": 3029, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3029, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_862_0", + "timepoint": 862, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 862, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_782_0", + "timepoint": 782, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 782, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_334_0", + "timepoint": 334, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 334, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_201_0", + "timepoint": 201, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 201, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; curette,patient,touching; head surgeon,forceps,holding; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,curette,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_794_0", + "timepoint": 794, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 794, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding; circulator,microscope,closeto; patient,operating table,lyingon; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1916_0", + "timepoint": 1916, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1916, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,needle,applying; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_690_0", + "timepoint": 690, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 690, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2749_0", + "timepoint": 2749, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2749, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_23_0", + "timepoint": 23, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 23, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_2502_0", + "timepoint": 2502, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2502, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "MISS_1_1_749_0", + "timepoint": 749, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 749, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,closeto; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_116_0", + "timepoint": 116, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 116, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_802_0", + "timepoint": 802, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 802, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,antiseptic,holding; assistant,forceps,holding; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_268_0", + "timepoint": 268, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 268, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,looking; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_3000_0", + "timepoint": 3000, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3000, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,instrument table,looking; assistant,syringe,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1789_0", + "timepoint": 1789, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1789, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; needle,patient,touching; assistant,instrument table,closeto; assistant,syringe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3040_0", + "timepoint": 3040, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3040, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,tissue paper,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_48_0", + "timepoint": 48, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 48, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_158_0", + "timepoint": 158, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 158, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound machine,looking; patient,operating table,lyingon; assistant,operating table,closeto; assistant,patient,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_318_0", + "timepoint": 318, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 318, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_749_0", + "timepoint": 749, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 749, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1480_0", + "timepoint": 1480, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1480, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2280_0", + "timepoint": 2280, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2280, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,needle,holding; assistant,syringe,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1937_0", + "timepoint": 1937, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1937, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,needle,holding; syringe,needle,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_2_2_213_0", + "timepoint": 213, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 213, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,health monitor,looking; forceps,cotton,touching; head surgeon,scalpel,holding; scalpel,patient,touching; head surgeon,patient,cutting; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1856_0", + "timepoint": 1856, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1856, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,patient,disinfection; assistant,antiseptic,applying; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_116_0", + "timepoint": 116, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 116, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto; circulator,syringe,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_629_0", + "timepoint": 629, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 629, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1453_0", + "timepoint": 1453, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1453, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,needle,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_441_0", + "timepoint": 441, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 441, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating room,entering; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_175_0", + "timepoint": 175, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 175, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1962_0", + "timepoint": 1962, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1962, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; syringe,needle,touching; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_712_0", + "timepoint": 712, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 712, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,curette,holding; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1020_0", + "timepoint": 1020, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1020, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1447_0", + "timepoint": 1447, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1447, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2281_0", + "timepoint": 2281, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2281, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,operating table,closeto; assistant,operating table,closeto; syringe,needle,touching; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_67_0", + "timepoint": 67, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 67, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1973_0", + "timepoint": 1973, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1973, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1472_0", + "timepoint": 1472, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1472, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_2_2_340_0", + "timepoint": 340, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 340, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; forceps,cotton,touching; cotton,patient,touching; head surgeon,scalpel,holding; anesthetist,health monitor,looking; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,aspirating; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_1_1_1614_0", + "timepoint": 1614, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1614, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_680_0", + "timepoint": 680, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 680, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_460_0", + "timepoint": 460, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 460, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_844_0", + "timepoint": 844, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 844, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_493_0", + "timepoint": 493, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 493, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; scalpel,patient,touching; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; curette,patient,touching; head surgeon,herbal disk,removing; patient,operating table,lyingon; head surgeon,curette,holding; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2018_0", + "timepoint": 2018, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2018, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,needle,inserting; assistant,syringe,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1832_0", + "timepoint": 1832, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1832, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,antiseptic,applying; assistant,patient,disinfection; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_572_0", + "timepoint": 572, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 572, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_762_0", + "timepoint": 762, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 762, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1659_0", + "timepoint": 1659, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1659, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue mark,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,needle,inserting; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1961_0", + "timepoint": 1961, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1961, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,needle,inserting; assistant,syringe,holding; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1098_0", + "timepoint": 1098, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1098, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; forceps,cotton,touching; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2730_0", + "timepoint": 2730, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2730, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,needle,holding; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1582_0", + "timepoint": 1582, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1582, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1737_0", + "timepoint": 1737, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1737, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,tissue mark,looking; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1455_0", + "timepoint": 1455, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1455, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_51_0", + "timepoint": 51, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 51, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,scalpel,looking; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_2_2_273_0", + "timepoint": 273, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 273, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,scalpel,holding; assistant,forceps,holding; forceps,cotton,touching; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_462_0", + "timepoint": 462, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 462, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering; circulator,operating table,manipulating; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "MISS_3_6_40_0", + "timepoint": 40, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 40, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,curette,holding; patient,operating table,lyingon; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "MISS_3_6_55_0", + "timepoint": 55, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 55, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,curette,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2611_0", + "timepoint": 2611, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2611, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_248_0", + "timepoint": 248, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 248, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_10_0", + "timepoint": 10, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 10, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1158_0", + "timepoint": 1158, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1158, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,inserting; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1111_0", + "timepoint": 1111, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1111, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,body marker,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2802_0", + "timepoint": 2802, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2802, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,holding; patient,operating table,lyingon; syringe,needle,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2046_0", + "timepoint": 2046, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2046, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; needle,patient,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_3_4_202_0", + "timepoint": 202, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 202, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_347_0", + "timepoint": 347, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 347, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "MISS_1_1_2115_0", + "timepoint": 2115, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2115, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,patient,disinfection; cotton,patient,touching; assistant,forceps,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2393_0", + "timepoint": 2393, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2393, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,ultrasound probe,controlling; syringe,needle,touching; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2414_0", + "timepoint": 2414, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2414, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound machine,looking; needle,patient,touching; assistant,needle,removing; assistant,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_3_6_412_0", + "timepoint": 412, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 412, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,forceps,holding; head surgeon,herbal disk,removing; assistant,instrument table,looking; head surgeon,microscope eye,looking; assistant,forceps,holding; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; curette,patient,touching; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,curette,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_602_0", + "timepoint": 602, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 602, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2023_0", + "timepoint": 2023, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2023, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; needle,patient,touching; assistant,needle,inserting; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,syringe,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2167_0", + "timepoint": 2167, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2167, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,operating table,closeto; assistant,needle,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; syringe,needle,touching; assistant,syringe,injecting; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_869_0", + "timepoint": 869, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 869, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2383_0", + "timepoint": 2383, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2383, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1522_0", + "timepoint": 1522, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1522, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,patient,anaesthetising; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,microscope controller,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_1350_0", + "timepoint": 1350, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1350, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,body marker,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2567_0", + "timepoint": 2567, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2567, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; assistant,needle,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_1_1_2534_0", + "timepoint": 2534, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2534, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,patient,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_107_0", + "timepoint": 107, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 107, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_7_0", + "timepoint": 7, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 7, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_1780_0", + "timepoint": 1780, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1780, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,antiseptic,holding; assistant,instrument table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1053_0", + "timepoint": 1053, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1053, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1586_0", + "timepoint": 1586, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1586, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_385_0", + "timepoint": 385, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 385, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; patient,operating table,touching; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_95_0", + "timepoint": 95, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 95, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1440_0", + "timepoint": 1440, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1440, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_608_0", + "timepoint": 608, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 608, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_483_0", + "timepoint": 483, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 483, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1483_0", + "timepoint": 1483, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1483, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,inserting; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_592_0", + "timepoint": 592, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 592, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_948_0", + "timepoint": 948, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 948, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,antiseptic,holding; assistant,patient,disinfection; cotton,patient,touching; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1052_0", + "timepoint": 1052, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1052, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; assistant,forceps,holding; assistant,patient,aspirating; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1808_0", + "timepoint": 1808, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1808, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,ultrasound probe,controlling; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2829_0", + "timepoint": 2829, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2829, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,removing; assistant,operating table,closeto; head surgeon,ultrasound machine,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_554_0", + "timepoint": 554, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 554, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_572_0", + "timepoint": 572, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 572, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_829_0", + "timepoint": 829, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 829, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_1617_0", + "timepoint": 1617, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1617, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,tissue mark,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,needle,inserting; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_16_0", + "timepoint": 16, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 16, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound machine,looking; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2701_0", + "timepoint": 2701, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2701, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,operating table,closeto; assistant,needle,holding; assistant,syringe,injecting; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_942_0", + "timepoint": 942, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 942, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2520_0", + "timepoint": 2520, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2520, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,tissue paper,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_310_0", + "timepoint": 310, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 310, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; forceps,patient,touching; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,health monitor,looking; head surgeon,forceps,holding; curette,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "Ultrasound_3_4_1159_0", + "timepoint": 1159, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1159, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,body marker,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_1694_0", + "timepoint": 1694, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1694, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_998_0", + "timepoint": 998, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 998, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,microscope,manipulating; patient,operating table,lyingon; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_894_0", + "timepoint": 894, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 894, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; scissors,cotton,touching; circulator,ultrasound machine,closeto; assistant,operating table,closeto; patient,operating table,lyingon; circulator,ultrasound machine,manipulating; assistant,scissors,holding; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2196_0", + "timepoint": 2196, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2196, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,needle,holding; assistant,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,syringe,injecting; syringe,needle,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1046_0", + "timepoint": 1046, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1046, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1573_0", + "timepoint": 1573, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1573, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,needle,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_2_2_6_0", + "timepoint": 6, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 6, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,patient,manipulating; anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1998_0", + "timepoint": 1998, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1998, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; assistant,instrument table,looking; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_4_5_2209_0", + "timepoint": 2209, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2209, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,needle,holding; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_2098_0", + "timepoint": 2098, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2098, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; assistant,forceps,holding; assistant,operating table,closeto; anesthetist,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_512_0", + "timepoint": 512, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 512, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; circulator,instrument table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,controlling; head surgeon,operating table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1629_0", + "timepoint": 1629, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1629, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,needle,inserting; needle,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_588_0", + "timepoint": 588, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 588, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1901_0", + "timepoint": 1901, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1901, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,operating table,closeto; ultrasound probe,patient,touching; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_1_1_350_0", + "timepoint": 350, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 350, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating room,entering; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_549_0", + "timepoint": 549, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 549, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1208_0", + "timepoint": 1208, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1208, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,body marker,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1737_0", + "timepoint": 1737, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1737, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,needle,holding; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking " + } + ] + }, + { + "id": "MISS_1_1_1677_0", + "timepoint": 1677, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1677, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2207_0", + "timepoint": 2207, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2207, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; needle,patient,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; assistant,operating table,closeto; syringe,needle,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2853_0", + "timepoint": 2853, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2853, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,syringe,removing; ultrasound probe,patient,touching; syringe,needle,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_867_0", + "timepoint": 867, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 867, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2328_0", + "timepoint": 2328, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2328, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_2433_0", + "timepoint": 2433, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2433, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,patient,manipulating; head surgeon,operating table,closeto; assistant,patient,manipulating; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_447_0", + "timepoint": 447, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 447, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_997_0", + "timepoint": 997, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 997, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,body marker,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_558_0", + "timepoint": 558, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 558, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating room,entering; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_93_0", + "timepoint": 93, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 93, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; assistant,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding " + } + ] + }, + { + "id": "MISS_3_6_1028_0", + "timepoint": 1028, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1028, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,patient,aspirating; patient,operating table,lyingon; assistant,operating table,closeto; cotton,patient,touching; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,health monitor,looking; forceps,cotton,touching; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1995_0", + "timepoint": 1995, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1995, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; syringe,needle,touching; assistant,syringe,injecting; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_25_0", + "timepoint": 25, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 25, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering; circulator,operating table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_1_1_5_0", + "timepoint": 5, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 5, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating room,entering " + } + ] + }, + { + "id": "MISS_3_6_554_0", + "timepoint": 554, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 554, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,scalpel,holding; head surgeon,herbal disk,removing; head surgeon,microscope eye,looking; head surgeon,curette,holding; scalpel,patient,touching; curette,patient,touching; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1503_0", + "timepoint": 1503, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1503, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_230_0", + "timepoint": 230, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 230, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2243_0", + "timepoint": 2243, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2243, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; patient,operating table,lyingon; syringe,needle,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_122_0", + "timepoint": 122, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 122, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound gel,applying; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1877_0", + "timepoint": 1877, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1877, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2372_0", + "timepoint": 2372, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2372, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,needle,holding; assistant,syringe,injecting; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_136_0", + "timepoint": 136, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 136, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1220_0", + "timepoint": 1220, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1220, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,inserting; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "MISS_1_1_308_0", + "timepoint": 308, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 308, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating room,entering; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_743_0", + "timepoint": 743, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 743, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_2016_0", + "timepoint": 2016, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2016, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,patient,disinfection; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_15_0", + "timepoint": 15, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 15, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound machine,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1119_0", + "timepoint": 1119, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1119, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,needle,inserting; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_196_0", + "timepoint": 196, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 196, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; curette,patient,touching; head surgeon,microscope eye,looking; head surgeon,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_67_0", + "timepoint": 67, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 67, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_765_0", + "timepoint": 765, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 765, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,microscope,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2883_0", + "timepoint": 2883, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2883, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,looking; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2351_0", + "timepoint": 2351, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2351, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,syringe,holding; assistant,needle,holding; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2178_0", + "timepoint": 2178, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2178, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; syringe,needle,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1120_0", + "timepoint": 1120, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1120, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,body marker,holding; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_633_0", + "timepoint": 633, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 633, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2096_0", + "timepoint": 2096, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2096, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,needle,holding; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_552_0", + "timepoint": 552, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 552, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,curette,holding; curette,patient,touching; head surgeon,scalpel,holding; head surgeon,herbal disk,removing; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; scalpel,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_760_0", + "timepoint": 760, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 760, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_380_0", + "timepoint": 380, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 380, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; forceps,cotton,touching; assistant,forceps,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,patient,aspirating; head surgeon,scalpel,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1054_0", + "timepoint": 1054, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1054, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,antiseptic,holding; patient,operating table,lyingon; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2369_0", + "timepoint": 2369, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2369, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,needle,removing; assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_348_0", + "timepoint": 348, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 348, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; head surgeon,scalpel,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; assistant,patient,aspirating; forceps,cotton,touching; anesthetist,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_899_0", + "timepoint": 899, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 899, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "MISS_1_1_2084_0", + "timepoint": 2084, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2084, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_3_4_1027_0", + "timepoint": 1027, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1027, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,body marker,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_1007_0", + "timepoint": 1007, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1007, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,aspirating; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,health monitor,looking; forceps,cotton,touching; cotton,patient,touching; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1263_0", + "timepoint": 1263, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1263, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,patient,manipulating; anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_692_0", + "timepoint": 692, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 692, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_77_0", + "timepoint": 77, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 77, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,curette,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1243_0", + "timepoint": 1243, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1243, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_814_0", + "timepoint": 814, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 814, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,forceps,holding; patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1990_0", + "timepoint": 1990, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1990, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; ultrasound probe,patient,touching; assistant,ultrasound probe,looking; patient,operating table,lyingon; assistant,needle,inserting; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2522_0", + "timepoint": 2522, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2522, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,needle,holding; assistant,syringe,injecting; patient,operating table,lyingon; syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_863_0", + "timepoint": 863, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 863, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_179_0", + "timepoint": 179, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 179, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_313_0", + "timepoint": 313, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 313, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_234_0", + "timepoint": 234, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 234, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; head surgeon,herbal disk,removing; assistant,operating table,closeto; head surgeon,curette,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,forceps,holding; head surgeon,microscope eye,looking; curette,patient,touching; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2533_0", + "timepoint": 2533, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2533, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,holding; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2113_0", + "timepoint": 2113, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2113, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,operating table,closeto; patient,operating table,lyingon; syringe,needle,touching; needle,patient,touching; assistant,needle,holding; head surgeon,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_976_0", + "timepoint": 976, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 976, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,body marker,holding " + } + ] + }, + { + "id": "MISS_1_1_239_0", + "timepoint": 239, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 239, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_250_0", + "timepoint": 250, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 250, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1086_0", + "timepoint": 1086, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1086, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_501_0", + "timepoint": 501, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 501, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,scalpel,holding; head surgeon,curette,holding; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; curette,patient,touching; assistant,operating table,closeto; head surgeon,microscope eye,looking; scalpel,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1663_0", + "timepoint": 1663, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1663, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; assistant,needle,inserting " + } + ] + }, + { + "id": "MISS_1_1_2485_0", + "timepoint": 2485, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2485, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1069_0", + "timepoint": 1069, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1069, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,aspirating; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,health monitor,looking; forceps,cotton,touching; cotton,patient,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2741_0", + "timepoint": 2741, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2741, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_298_0", + "timepoint": 298, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 298, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "MISS_2_2_520_0", + "timepoint": 520, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 520, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; assistant,instrument table,closeto; circulator,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,microscope controller,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1335_0", + "timepoint": 1335, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1335, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_616_0", + "timepoint": 616, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 616, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_590_0", + "timepoint": 590, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 590, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_1111_0", + "timepoint": 1111, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1111, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,manipulating; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,antiseptic,holding; circulator,ultrasound machine,closeto; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1058_0", + "timepoint": 1058, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1058, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_1148_0", + "timepoint": 1148, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1148, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,operating table,closeto; assistant,forceps,holding; circulator,ultrasound machine,closeto; patient,operating table,lyingon; circulator,ultrasound machine,manipulating; assistant,antiseptic,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2546_0", + "timepoint": 2546, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2546, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,syringe,injecting; syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_251_0", + "timepoint": 251, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 251, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2077_0", + "timepoint": 2077, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2077, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; forceps,cotton,touching; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_1_1_796_0", + "timepoint": 796, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 796, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,microscope,closeto; anesthetist,anesthesia equipment,holding; circulator,microscope,manipulating " + } + ] + }, + { + "id": "MISS_1_1_394_0", + "timepoint": 394, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 394, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; anesthetist,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_2690_0", + "timepoint": 2690, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2690, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,needle,holding; syringe,needle,touching; needle,patient,touching; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_2662_0", + "timepoint": 2662, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2662, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2716_0", + "timepoint": 2716, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2716, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; patient,operating table,lyingon; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_606_0", + "timepoint": 606, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 606, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1740_0", + "timepoint": 1740, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1740, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue mark,looking; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,needle,inserting; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_178_0", + "timepoint": 178, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 178, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1650_0", + "timepoint": 1650, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1650, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; assistant,needle,inserting; patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2304_0", + "timepoint": 2304, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2304, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; syringe,needle,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2450_0", + "timepoint": 2450, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2450, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_634_0", + "timepoint": 634, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 634, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2540_0", + "timepoint": 2540, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2540, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_230_0", + "timepoint": 230, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 230, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1237_0", + "timepoint": 1237, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1237, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,body marker,applying; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_498_0", + "timepoint": 498, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 498, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2942_0", + "timepoint": 2942, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2942, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; tissue paper,patient,touching; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "MISS_1_1_1275_0", + "timepoint": 1275, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1275, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,patient,manipulating; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_613_0", + "timepoint": 613, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 613, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1220_0", + "timepoint": 1220, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1220, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1732_0", + "timepoint": 1732, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1732, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,instrument table,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_770_0", + "timepoint": 770, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 770, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1581_0", + "timepoint": 1581, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1581, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2469_0", + "timepoint": 2469, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2469, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,needle,holding; head surgeon,vertebrae,scanning; syringe,needle,touching; head surgeon,ultrasound screen,looking; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_2_2_1201_0", + "timepoint": 1201, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1201, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; circulator,ultrasound machine,manipulating; assistant,antiseptic,holding; circulator,ultrasound machine,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_1529_0", + "timepoint": 1529, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1529, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,manipulating; anesthetist,patient,anaesthetising; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_965_0", + "timepoint": 965, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 965, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; cotton,patient,touching; assistant,patient,disinfection; forceps,cotton,touching; patient,operating table,lyingon; assistant,antiseptic,holding; assistant,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_511_0", + "timepoint": 511, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 511, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_331_0", + "timepoint": 331, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 331, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,ultrasound gel,applying " + } + ] + }, + { + "id": "MISS_2_2_205_0", + "timepoint": 205, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 205, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; forceps,cotton,touching; scalpel,patient,touching; assistant,forceps,holding; anesthetist,operating table,closeto; anesthetist,health monitor,looking; head surgeon,patient,cutting; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2070_0", + "timepoint": 2070, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2070, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,syringe,injecting; assistant,needle,holding; assistant,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2652_0", + "timepoint": 2652, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2652, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; needle,patient,touching; assistant,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_2008_0", + "timepoint": 2008, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2008, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,patient,disinfection; anesthetist,operating table,closeto; assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_652_0", + "timepoint": 652, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 652, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_1341_0", + "timepoint": 1341, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1341, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; assistant,antiseptic,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2587_0", + "timepoint": 2587, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2587, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; patient,operating table,lyingon; assistant,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_835_0", + "timepoint": 835, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 835, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2028_0", + "timepoint": 2028, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2028, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; assistant,syringe,holding " + } + ] + }, + { + "id": "MISS_1_1_343_0", + "timepoint": 343, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 343, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_1105_0", + "timepoint": 1105, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1105, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_2_2_1062_0", + "timepoint": 1062, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1062, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,antiseptic,holding; circulator,ultrasound machine,looking; assistant,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_168_0", + "timepoint": 168, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 168, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,herbal disk,removing; curette,patient,touching; patient,operating table,lyingon; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_608_0", + "timepoint": 608, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 608, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_986_0", + "timepoint": 986, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 986, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,microscope,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_203_0", + "timepoint": 203, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 203, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_2303_0", + "timepoint": 2303, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2303, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,instrument table,looking; assistant,instrument table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1593_0", + "timepoint": 1593, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1593, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,inserting; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2542_0", + "timepoint": 2542, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2542, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "MISS_1_1_1650_0", + "timepoint": 1650, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1650, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1450_0", + "timepoint": 1450, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1450, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2558_0", + "timepoint": 2558, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2558, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; syringe,needle,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2051_0", + "timepoint": 2051, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2051, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_3_6_206_0", + "timepoint": 206, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 206, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,forceps,holding; head surgeon,curette,holding; head surgeon,herbal disk,removing; curette,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "MISS_3_6_16_0", + "timepoint": 16, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 16, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_769_0", + "timepoint": 769, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 769, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; circulator,microscope,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1343_0", + "timepoint": 1343, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1343, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,patient,anaesthetising; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_571_0", + "timepoint": 571, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 571, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_579_0", + "timepoint": 579, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 579, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1009_0", + "timepoint": 1009, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1009, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1190_0", + "timepoint": 1190, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1190, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,ultrasound machine,closeto; circulator,ultrasound machine,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_3063_0", + "timepoint": 3063, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3063, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,tissue paper,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2208_0", + "timepoint": 2208, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2208, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,needle,holding; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,holding; patient,operating table,lyingon; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1134_0", + "timepoint": 1134, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1134, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,operating table,closeto; circulator,ultrasound machine,manipulating; assistant,forceps,holding; assistant,antiseptic,holding; patient,operating table,lyingon; assistant,operating table,closeto; circulator,ultrasound machine,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3007_0", + "timepoint": 3007, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3007, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,applying; tissue paper,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1163_0", + "timepoint": 1163, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1163, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound probe,looking; ultrasound probe,patient,touching; assistant,needle,inserting; assistant,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2397_0", + "timepoint": 2397, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2397, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,syringe,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_752_0", + "timepoint": 752, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 752, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_315_0", + "timepoint": 315, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 315, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering; anesthetist,operating table,manipulating; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_1149_0", + "timepoint": 1149, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1149, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; forceps,cotton,touching; assistant,operating table,closeto; assistant,antiseptic,holding; head surgeon,operating table,closeto; circulator,ultrasound machine,manipulating; patient,operating table,lyingon; circulator,ultrasound machine,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2186_0", + "timepoint": 2186, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2186, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,ultrasound screen,looking; assistant,syringe,injecting; head surgeon,needle,holding; needle,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1134_0", + "timepoint": 1134, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1134, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; assistant,ultrasound probe,looking; ultrasound probe,patient,touching; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_5_6_2348_0", + "timepoint": 2348, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2348, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,syringe,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_2_2_472_0", + "timepoint": 472, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 472, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; assistant,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1564_0", + "timepoint": 1564, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1564, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; needle,patient,touching; ultrasound probe,patient,touching; assistant,needle,inserting; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_795_0", + "timepoint": 795, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 795, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching; assistant,antiseptic,holding; assistant,instrument table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_3_6_481_0", + "timepoint": 481, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 481, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,forceps,holding; assistant,operating table,closeto; head surgeon,scalpel,holding; scalpel,patient,touching; curette,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "Ultrasound_5_6_1055_0", + "timepoint": 1055, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1055, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1643_0", + "timepoint": 1643, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1643, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,instrument table,closeto; head surgeon,ultrasound screen,looking; assistant,instrument table,looking; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1572_0", + "timepoint": 1572, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1572, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2201_0", + "timepoint": 2201, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2201, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; head surgeon,needle,holding; syringe,needle,touching; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1670_0", + "timepoint": 1670, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1670, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,tissue mark,looking; assistant,needle,inserting; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1640_0", + "timepoint": 1640, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1640, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; assistant,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_1366_0", + "timepoint": 1366, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1366, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; circulator,operating table,looking; anesthetist,patient,anaesthetising; assistant,instrument table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "MISS_3_6_1149_0", + "timepoint": 1149, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1149, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,health monitor,looking; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_285_0", + "timepoint": 285, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 285, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "MISS_3_6_156_0", + "timepoint": 156, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 156, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; patient,operating table,lyingon; curette,patient,touching; assistant,forceps,holding; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; head surgeon,curette,holding; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "Ultrasound_4_5_1988_0", + "timepoint": 1988, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1988, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; needle,patient,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_4_5_1693_0", + "timepoint": 1693, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1693, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_4_5_753_0", + "timepoint": 753, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 753, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1667_0", + "timepoint": 1667, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1667, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; assistant,needle,inserting; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_266_0", + "timepoint": 266, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 266, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_715_0", + "timepoint": 715, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 715, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1494_0", + "timepoint": 1494, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1494, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_397_0", + "timepoint": 397, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 397, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_1841_0", + "timepoint": 1841, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1841, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,syringe,injecting; needle,patient,touching; syringe,needle,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2202_0", + "timepoint": 2202, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2202, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,syringe,injecting; patient,operating table,lyingon; syringe,needle,touching; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2927_0", + "timepoint": 2927, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2927, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,syringe,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_391_0", + "timepoint": 391, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 391, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_817_0", + "timepoint": 817, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 817, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,antiseptic,holding; assistant,operating table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1561_0", + "timepoint": 1561, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1561, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_961_0", + "timepoint": 961, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 961, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,microscope,manipulating; anesthetist,microscope,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_221_0", + "timepoint": 221, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 221, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1272_0", + "timepoint": 1272, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1272, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2266_0", + "timepoint": 2266, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2266, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,ultrasound screen,looking; syringe,needle,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,holding; needle,patient,touching; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_593_0", + "timepoint": 593, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 593, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_763_0", + "timepoint": 763, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 763, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1521_0", + "timepoint": 1521, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1521, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2009_0", + "timepoint": 2009, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2009, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,patient,disinfection; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_78_0", + "timepoint": 78, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 78, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking; circulator,antiseptic,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1204_0", + "timepoint": 1204, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1204, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_748_0", + "timepoint": 748, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 748, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,microscope screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_671_0", + "timepoint": 671, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 671, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,controlling; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,microscope screen,looking; assistant,microscope screen,looking " + } + ] + }, + { + "id": "MISS_3_6_1102_0", + "timepoint": 1102, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1102, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,microscope eye,looking; forceps,cotton,touching; assistant,forceps,holding; anesthetist,health monitor,looking; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_745_0", + "timepoint": 745, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 745, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "MISS_2_2_210_0", + "timepoint": 210, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 210, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; head surgeon,patient,cutting; forceps,cotton,touching; scalpel,patient,touching; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,scalpel,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_78_0", + "timepoint": 78, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 78, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound machine,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_233_0", + "timepoint": 233, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 233, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_541_0", + "timepoint": 541, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 541, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,assistant,closeto; patient,operating table,lyingon; assistant,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_691_0", + "timepoint": 691, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 691, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_29_0", + "timepoint": 29, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 29, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound machine,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_972_0", + "timepoint": 972, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 972, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; cotton,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; forceps,cotton,touching; head surgeon,microscope eye,looking; assistant,forceps,holding; assistant,patient,aspirating " + } + ] + }, + { + "id": "Ultrasound_3_4_2305_0", + "timepoint": 2305, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2305, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,operating table,closeto; needle,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_576_0", + "timepoint": 576, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 576, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,microscope controller,controlling; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_164_0", + "timepoint": 164, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 164, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; head surgeon,herbal disk,removing; assistant,forceps,holding; assistant,operating table,closeto; curette,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1105_0", + "timepoint": 1105, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1105, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,cotton,holding; anesthetist,health monitor,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2007_0", + "timepoint": 2007, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2007, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,syringe,injecting; head surgeon,needle,holding; patient,operating table,lyingon; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_3_6_121_0", + "timepoint": 121, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 121, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_964_0", + "timepoint": 964, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 964, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_786_0", + "timepoint": 786, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 786, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_89_0", + "timepoint": 89, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 89, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,antiseptic,looking; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_1325_0", + "timepoint": 1325, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1325, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,forceps,holding; assistant,operating table,closeto; assistant,antiseptic,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_837_0", + "timepoint": 837, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 837, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3277_0", + "timepoint": 3277, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3277, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; tissue paper,patient,touching; assistant,tissue paper,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1931_0", + "timepoint": 1931, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1931, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,syringe,holding; assistant,ultrasound probe,looking; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_253_0", + "timepoint": 253, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 253, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_713_0", + "timepoint": 713, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 713, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1170_0", + "timepoint": 1170, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1170, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating room,entering; assistant,operating room,entering; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_485_0", + "timepoint": 485, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 485, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1951_0", + "timepoint": 1951, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1951, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,syringe,holding; assistant,needle,inserting; patient,operating table,lyingon; assistant,ultrasound probe,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_33_0", + "timepoint": 33, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 33, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_598_0", + "timepoint": 598, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 598, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_1237_0", + "timepoint": 1237, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1237, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto; anesthetist,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_905_0", + "timepoint": 905, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 905, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2795_0", + "timepoint": 2795, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2795, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_35_0", + "timepoint": 35, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 35, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1476_0", + "timepoint": 1476, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1476, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_482_0", + "timepoint": 482, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 482, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_2289_0", + "timepoint": 2289, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2289, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; anesthetist,operating table,closeto; assistant,instrument table,looking; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2697_0", + "timepoint": 2697, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2697, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,syringe,injecting; syringe,needle,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1500_0", + "timepoint": 1500, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1500, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; assistant,needle,inserting; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_429_0", + "timepoint": 429, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 429, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; assistant,instrument table,looking; forceps,cotton,touching; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; assistant,instrument table,closeto; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1642_0", + "timepoint": 1642, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1642, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_2125_0", + "timepoint": 2125, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2125, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; cotton,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,patient,disinfection; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2831_0", + "timepoint": 2831, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2831, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,holding; head surgeon,ultrasound machine,looking; assistant,needle,removing; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_210_0", + "timepoint": 210, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 210, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1645_0", + "timepoint": 1645, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1645, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,tissue mark,looking; assistant,needle,inserting; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_458_0", + "timepoint": 458, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 458, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2175_0", + "timepoint": 2175, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2175, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,patient,disinfection; assistant,forceps,holding; head surgeon,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; assistant,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_834_0", + "timepoint": 834, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 834, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_934_0", + "timepoint": 934, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 934, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_488_0", + "timepoint": 488, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 488, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon; circulator,operating table,manipulating; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_716_0", + "timepoint": 716, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 716, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,patient,manipulating; head surgeon,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_3089_0", + "timepoint": 3089, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3089, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,tissue paper,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_743_0", + "timepoint": 743, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 743, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,microscope screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2753_0", + "timepoint": 2753, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2753, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,needle,holding; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2612_0", + "timepoint": 2612, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2612, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound screen,looking; syringe,needle,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; needle,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_455_0", + "timepoint": 455, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 455, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_225_0", + "timepoint": 225, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 225, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1129_0", + "timepoint": 1129, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1129, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,microscope eye,looking; anesthetist,health monitor,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_786_0", + "timepoint": 786, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 786, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,instrument table,closeto; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_730_0", + "timepoint": 730, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 730, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_1035_0", + "timepoint": 1035, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1035, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,microscope,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2653_0", + "timepoint": 2653, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2653, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; syringe,needle,touching; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_801_0", + "timepoint": 801, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 801, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; assistant,instrument table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_456_0", + "timepoint": 456, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 456, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2205_0", + "timepoint": 2205, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2205, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_657_0", + "timepoint": 657, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 657, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2349_0", + "timepoint": 2349, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2349, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2252_0", + "timepoint": 2252, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2252, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,syringe,removing; ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_1_1_1132_0", + "timepoint": 1132, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1132, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_374_0", + "timepoint": 374, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 374, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,instrument table,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2955_0", + "timepoint": 2955, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2955, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_3085_0", + "timepoint": 3085, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3085, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,tissue paper,holding " + } + ] + }, + { + "id": "MISS_2_2_69_0", + "timepoint": 69, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 69, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,scalpel,holding; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; anesthetist,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1203_0", + "timepoint": 1203, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1203, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3011_0", + "timepoint": 3011, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3011, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,instrument table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_590_0", + "timepoint": 590, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 590, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,curette,holding; head surgeon,herbal disk,removing; scalpel,patient,touching; head surgeon,microscope eye,looking; curette,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1841_0", + "timepoint": 1841, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1841, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,syringe,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_675_0", + "timepoint": 675, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 675, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_990_0", + "timepoint": 990, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 990, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,antiseptic,holding; assistant,patient,disinfection; assistant,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; cotton,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_539_0", + "timepoint": 539, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 539, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope controller,controlling; anesthetist,anesthesia equipment,controlling; circulator,instrument table,closeto; head surgeon,operating table,closeto; anesthetist,health monitor,looking; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2273_0", + "timepoint": 2273, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2273, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; needle,patient,touching; syringe,needle,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1580_0", + "timepoint": 1580, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1580, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; assistant,ultrasound probe,looking; patient,operating table,lyingon; assistant,needle,inserting; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2044_0", + "timepoint": 2044, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2044, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,syringe,injecting; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_366_0", + "timepoint": 366, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 366, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1049_0", + "timepoint": 1049, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1049, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,body marker,holding; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1297_0", + "timepoint": 1297, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1297, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_764_0", + "timepoint": 764, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 764, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_630_0", + "timepoint": 630, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 630, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_1223_0", + "timepoint": 1223, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1223, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; patient,operating table,lyingon; assistant,operating table,closeto; assistant,forceps,holding; circulator,ultrasound machine,closeto; assistant,antiseptic,holding; head surgeon,operating table,closeto; circulator,ultrasound machine,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1428_0", + "timepoint": 1428, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1428, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2254_0", + "timepoint": 2254, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2254, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1152_0", + "timepoint": 1152, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1152, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,body marker,holding; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1724_0", + "timepoint": 1724, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1724, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; assistant,syringe,holding " + } + ] + }, + { + "id": "MISS_1_1_1706_0", + "timepoint": 1706, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1706, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,looking; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1456_0", + "timepoint": 1456, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1456, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1148_0", + "timepoint": 1148, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1148, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating room,entering; assistant,operating room,entering; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1635_0", + "timepoint": 1635, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1635, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,tissue mark,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1276_0", + "timepoint": 1276, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1276, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,patient,manipulating; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_680_0", + "timepoint": 680, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 680, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1733_0", + "timepoint": 1733, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1733, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,needle,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_107_0", + "timepoint": 107, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 107, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,scalpel,holding; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1682_0", + "timepoint": 1682, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1682, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_533_0", + "timepoint": 533, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 533, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_642_0", + "timepoint": 642, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 642, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2877_0", + "timepoint": 2877, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2877, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound machine,looking; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,needle,holding; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1630_0", + "timepoint": 1630, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1630, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,tissue mark,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_4_5_146_0", + "timepoint": 146, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 146, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1952_0", + "timepoint": 1952, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1952, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; patient,operating table,lyingon; assistant,operating table,closeto; syringe,needle,touching; needle,patient,touching; assistant,syringe,injecting; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3263_0", + "timepoint": 3263, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3263, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; assistant,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_416_0", + "timepoint": 416, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 416, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; head surgeon,forceps,holding; assistant,forceps,holding; assistant,instrument table,looking; assistant,operating table,closeto; head surgeon,microscope eye,looking; curette,patient,touching; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1145_0", + "timepoint": 1145, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1145, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; assistant,ultrasound probe,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_942_0", + "timepoint": 942, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 942, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; patient,operating table,lyingon; assistant,forceps,holding; cotton,patient,touching; forceps,cotton,touching; assistant,patient,disinfection; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2667_0", + "timepoint": 2667, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2667, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2047_0", + "timepoint": 2047, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2047, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; syringe,needle,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2165_0", + "timepoint": 2165, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2165, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; syringe,needle,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1150_0", + "timepoint": 1150, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1150, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating room,entering; head surgeon,operating room,entering; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_968_0", + "timepoint": 968, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 968, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_518_0", + "timepoint": 518, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 518, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3243_0", + "timepoint": 3243, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3243, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,applying " + } + ] + }, + { + "id": "MISS_2_2_767_0", + "timepoint": 767, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 767, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope screen,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1217_0", + "timepoint": 1217, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1217, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1818_0", + "timepoint": 1818, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1818, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,inserting; patient,operating table,lyingon; assistant,operating table,closeto; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2849_0", + "timepoint": 2849, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2849, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,holding; assistant,operating table,closeto; assistant,instrument table,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_14_0", + "timepoint": 14, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 14, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2417_0", + "timepoint": 2417, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2417, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,ultrasound screen,looking; assistant,instrument table,closeto; assistant,syringe,holding; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1528_0", + "timepoint": 1528, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1528, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2571_0", + "timepoint": 2571, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2571, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_3_4_1804_0", + "timepoint": 1804, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1804, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,inserting; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2180_0", + "timepoint": 2180, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2180, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; syringe,needle,touching; patient,operating table,lyingon; assistant,needle,holding; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_2530_0", + "timepoint": 2530, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2530, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_827_0", + "timepoint": 827, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 827, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_434_0", + "timepoint": 434, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 434, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,touching; circulator,patient,closeto; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_237_0", + "timepoint": 237, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 237, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,instrument table,closeto; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1277_0", + "timepoint": 1277, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1277, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_79_0", + "timepoint": 79, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 79, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "MISS_2_2_713_0", + "timepoint": 713, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 713, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,microscope screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_940_0", + "timepoint": 940, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 940, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1387_0", + "timepoint": 1387, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1387, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_821_0", + "timepoint": 821, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 821, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,antiseptic,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_2569_0", + "timepoint": 2569, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2569, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,manipulating; anesthetist,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_267_0", + "timepoint": 267, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 267, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_696_0", + "timepoint": 696, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 696, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; circulator,ultrasound machine,looking; head surgeon,patient,manipulating; assistant,instrument table,looking; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_632_0", + "timepoint": 632, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 632, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; head surgeon,microscope eye,looking; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,scalpel,holding; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_901_0", + "timepoint": 901, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 901, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,ultrasound probe,looking; assistant,body marker,applying; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; body marker,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_350_0", + "timepoint": 350, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 350, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; cotton,patient,touching; head surgeon,scalpel,holding; assistant,patient,aspirating; head surgeon,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2323_0", + "timepoint": 2323, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2323, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; syringe,needle,touching; needle,patient,touching; ultrasound probe,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_2240_0", + "timepoint": 2240, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2240, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; cotton,patient,touching; patient,operating table,lyingon; assistant,forceps,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_343_0", + "timepoint": 343, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 343, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1067_0", + "timepoint": 1067, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1067, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_475_0", + "timepoint": 475, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 475, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_449_0", + "timepoint": 449, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 449, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; curette,patient,touching; assistant,scalpel,holding; head surgeon,curette,holding; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1434_0", + "timepoint": 1434, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1434, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1616_0", + "timepoint": 1616, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1616, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,tissue mark,looking; assistant,needle,inserting " + } + ] + }, + { + "id": "MISS_3_6_169_0", + "timepoint": 169, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 169, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,herbal disk,removing; assistant,forceps,holding; assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,curette,holding; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2456_0", + "timepoint": 2456, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2456, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,needle,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_971_0", + "timepoint": 971, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 971, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; assistant,forceps,holding; cotton,patient,touching; assistant,patient,disinfection " + } + ] + }, + { + "id": "MISS_1_1_1956_0", + "timepoint": 1956, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1956, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,patient,disinfection; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,antiseptic,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_325_0", + "timepoint": 325, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 325, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,ultrasound gel,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_2696_0", + "timepoint": 2696, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2696, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,vertebrae,scanning; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_293_0", + "timepoint": 293, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 293, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound gel,applying; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_1_1_788_0", + "timepoint": 788, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 788, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,closeto; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,microscope,manipulating " + } + ] + }, + { + "id": "MISS_2_2_248_0", + "timepoint": 248, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 248, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; scalpel,patient,touching; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,patient,cutting; assistant,forceps,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_548_0", + "timepoint": 548, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 548, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_291_0", + "timepoint": 291, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 291, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,looking; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_829_0", + "timepoint": 829, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 829, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_530_0", + "timepoint": 530, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 530, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_647_0", + "timepoint": 647, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 647, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_648_0", + "timepoint": 648, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 648, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1220_0", + "timepoint": 1220, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1220, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,antiseptic,holding; circulator,ultrasound machine,closeto; patient,operating table,lyingon; forceps,cotton,touching; circulator,ultrasound machine,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1024_0", + "timepoint": 1024, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1024, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1740_0", + "timepoint": 1740, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1740, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,holding; assistant,syringe,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_153_0", + "timepoint": 153, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 153, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,curette,holding; curette,patient,touching; anesthetist,operating table,closeto; assistant,forceps,holding; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2308_0", + "timepoint": 2308, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2308, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,instrument table,closeto; head surgeon,needle,holding; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1840_0", + "timepoint": 1840, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1840, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_248_0", + "timepoint": 248, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 248, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2860_0", + "timepoint": 2860, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2860, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,needle,holding; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; needle,patient,touching; syringe,needle,touching; assistant,syringe,removing " + } + ] + }, + { + "id": "Ultrasound_2_2_579_0", + "timepoint": 579, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 579, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_872_0", + "timepoint": 872, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 872, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; assistant,forceps,holding; assistant,patient,disinfection; assistant,antiseptic,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_2021_0", + "timepoint": 2021, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2021, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,inserting; needle,patient,touching; assistant,operating table,closeto; assistant,syringe,holding; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_568_0", + "timepoint": 568, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 568, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "MISS_1_1_155_0", + "timepoint": 155, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 155, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2095_0", + "timepoint": 2095, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2095, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; head surgeon,operating table,closeto; cotton,patient,touching; assistant,forceps,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1014_0", + "timepoint": 1014, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1014, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,body marker,holding; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1202_0", + "timepoint": 1202, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1202, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,body marker,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_443_0", + "timepoint": 443, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 443, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_710_0", + "timepoint": 710, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 710, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,microscope screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_401_0", + "timepoint": 401, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 401, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1232_0", + "timepoint": 1232, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1232, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_318_0", + "timepoint": 318, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 318, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_85_0", + "timepoint": 85, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 85, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,antiseptic,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_311_0", + "timepoint": 311, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 311, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,health monitor,looking; assistant,operating table,closeto; assistant,forceps,holding; head surgeon,scalpel,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,patient,aspirating; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_1029_0", + "timepoint": 1029, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1029, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,microscope,manipulating; anesthetist,operating table,closeto; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_649_0", + "timepoint": 649, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 649, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1156_0", + "timepoint": 1156, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1156, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1992_0", + "timepoint": 1992, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1992, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; head surgeon,needle,holding; assistant,syringe,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_193_0", + "timepoint": 193, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 193, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,ultrasound gel,applying; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3282_0", + "timepoint": 3282, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3282, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,tissue paper,holding " + } + ] + }, + { + "id": "MISS_1_1_2472_0", + "timepoint": 2472, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2472, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,manipulating " + } + ] + }, + { + "id": "MISS_2_2_261_0", + "timepoint": 261, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 261, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; patient,operating table,lyingon; anesthetist,operating table,closeto; forceps,cotton,touching; anesthetist,health monitor,looking; head surgeon,patient,cutting; assistant,operating table,closeto; scalpel,patient,touching; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "MISS_2_2_381_0", + "timepoint": 381, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 381, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching; cotton,patient,touching; assistant,operating table,closeto; assistant,patient,aspirating; head surgeon,scalpel,holding; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_3072_0", + "timepoint": 3072, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3072, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_901_0", + "timepoint": 901, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 901, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_937_0", + "timepoint": 937, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 937, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1101_0", + "timepoint": 1101, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1101, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; assistant,needle,inserting; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_547_0", + "timepoint": 547, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 547, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2097_0", + "timepoint": 2097, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2097, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,needle,holding; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,syringe,injecting; assistant,operating table,closeto; syringe,needle,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_208_0", + "timepoint": 208, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 208, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,ultrasound gel,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_851_0", + "timepoint": 851, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 851, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_127_0", + "timepoint": 127, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 127, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,ultrasound gel,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_864_0", + "timepoint": 864, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 864, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,microscope,manipulating; circulator,microscope,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_329_0", + "timepoint": 329, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 329, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,ultrasound gel,applying; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2649_0", + "timepoint": 2649, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2649, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,tissue paper,applying; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1200_0", + "timepoint": 1200, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1200, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2116_0", + "timepoint": 2116, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2116, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; syringe,needle,touching; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_338_0", + "timepoint": 338, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 338, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; patient,operating room,entering " + } + ] + }, + { + "id": "MISS_2_2_754_0", + "timepoint": 754, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 754, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,microscope screen,looking; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1116_0", + "timepoint": 1116, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1116, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,inserting; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1279_0", + "timepoint": 1279, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1279, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; assistant,needle,inserting; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1742_0", + "timepoint": 1742, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1742, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,needle,inserting; assistant,tissue mark,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_519_0", + "timepoint": 519, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 519, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating room,entering; circulator,assistant,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2130_0", + "timepoint": 2130, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2130, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,needle,holding; assistant,syringe,injecting; needle,patient,touching; syringe,needle,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_138_0", + "timepoint": 138, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 138, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1034_0", + "timepoint": 1034, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1034, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1132_0", + "timepoint": 1132, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1132, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2424_0", + "timepoint": 2424, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2424, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; patient,operating table,lyingon; head surgeon,patient,manipulating; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2015_0", + "timepoint": 2015, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2015, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching; assistant,needle,inserting; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,syringe,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1691_0", + "timepoint": 1691, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1691, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1769_0", + "timepoint": 1769, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1769, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "MISS_1_1_426_0", + "timepoint": 426, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 426, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,manipulating; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_982_0", + "timepoint": 982, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 982, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,body marker,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_534_0", + "timepoint": 534, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 534, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating room,entering; circulator,assistant,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1479_0", + "timepoint": 1479, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1479, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2990_0", + "timepoint": 2990, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2990, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,patient,scanning; patient,operating table,lyingon; assistant,instrument table,looking; assistant,operating table,closeto; assistant,syringe,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_932_0", + "timepoint": 932, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 932, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,body marker,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_1214_0", + "timepoint": 1214, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1214, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching; circulator,ultrasound machine,manipulating; assistant,antiseptic,holding; circulator,ultrasound machine,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_326_0", + "timepoint": 326, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 326, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound gel,applying; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3165_0", + "timepoint": 3165, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3165, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto; tissue paper,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_339_0", + "timepoint": 339, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 339, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_726_0", + "timepoint": 726, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 726, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_2_2_618_0", + "timepoint": 618, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 618, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,controlling; patient,operating table,lyingon; anesthetist,health monitor,looking; head surgeon,microscope screen,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,microscope screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1802_0", + "timepoint": 1802, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1802, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; assistant,needle,inserting; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2650_0", + "timepoint": 2650, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2650, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_2170_0", + "timepoint": 2170, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2170, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,needle,holding; syringe,needle,touching; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_363_0", + "timepoint": 363, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 363, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; head surgeon,microscope eye,looking; head surgeon,forceps,holding; forceps,patient,touching; head surgeon,herbal disk,removing; curette,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_231_0", + "timepoint": 231, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 231, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; anesthetist,operating table,closeto; head surgeon,forceps,holding; assistant,operating table,closeto; head surgeon,curette,holding; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; curette,patient,touching; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_256_0", + "timepoint": 256, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 256, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2270_0", + "timepoint": 2270, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2270, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_1131_0", + "timepoint": 1131, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1131, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,instrument table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1725_0", + "timepoint": 1725, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1725, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,tissue mark,looking; assistant,needle,inserting; needle,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1517_0", + "timepoint": 1517, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1517, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_325_0", + "timepoint": 325, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 325, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,aspirating; anesthetist,health monitor,looking; assistant,operating table,closeto; head surgeon,scalpel,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding; anesthetist,operating table,closeto; cotton,patient,touching; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_969_0", + "timepoint": 969, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 969, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,patient,disinfection; assistant,forceps,holding; assistant,antiseptic,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2200_0", + "timepoint": 2200, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2200, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,syringe,injecting; assistant,needle,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,needle,holding; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3138_0", + "timepoint": 3138, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3138, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; tissue paper,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1724_0", + "timepoint": 1724, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1724, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1589_0", + "timepoint": 1589, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1589, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_404_0", + "timepoint": 404, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 404, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; circulator,operating table,closeto; patient,operating table,touching " + } + ] + }, + { + "id": "MISS_1_1_1086_0", + "timepoint": 1086, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1086, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; circulator,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2341_0", + "timepoint": 2341, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2341, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1210_0", + "timepoint": 1210, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1210, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1627_0", + "timepoint": 1627, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1627, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_2_2_563_0", + "timepoint": 563, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 563, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,health monitor,looking; anesthetist,operating table,closeto; head surgeon,microscope controller,controlling; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1144_0", + "timepoint": 1144, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1144, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_834_0", + "timepoint": 834, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 834, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_520_0", + "timepoint": 520, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 520, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " scalpel,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; curette,patient,touching; head surgeon,herbal disk,removing; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,curette,holding; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3189_0", + "timepoint": 3189, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3189, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,tissue paper,applying; tissue paper,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2050_0", + "timepoint": 2050, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2050, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; assistant,operating table,closeto; assistant,needle,holding; syringe,needle,touching; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_635_0", + "timepoint": 635, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 635, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon; assistant,patient,manipulating " + } + ] + }, + { + "id": "MISS_1_1_2071_0", + "timepoint": 2071, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2071, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; forceps,cotton,touching; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_2_2_214_0", + "timepoint": 214, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 214, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling; circulator,syringe,looking " + } + ] + }, + { + "id": "MISS_1_1_607_0", + "timepoint": 607, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 607, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "MISS_2_2_517_0", + "timepoint": 517, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 517, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,microscope controller,controlling; head surgeon,operating table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_582_0", + "timepoint": 582, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 582, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1065_0", + "timepoint": 1065, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1065, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1284_0", + "timepoint": 1284, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1284, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,patient,manipulating; patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1508_0", + "timepoint": 1508, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1508, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1987_0", + "timepoint": 1987, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1987, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,ultrasound probe,looking; assistant,syringe,holding; needle,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_4_5_1319_0", + "timepoint": 1319, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1319, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,inserting; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_119_0", + "timepoint": 119, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 119, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_1_1_2029_0", + "timepoint": 2029, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2029, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,forceps,holding; anesthetist,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_3_4_1182_0", + "timepoint": 1182, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1182, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; assistant,body marker,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1464_0", + "timepoint": 1464, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1464, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_739_0", + "timepoint": 739, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 739, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1803_0", + "timepoint": 1803, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1803, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,inserting; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_3109_0", + "timepoint": 3109, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3109, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3002_0", + "timepoint": 3002, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3002, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_674_0", + "timepoint": 674, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 674, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1964_0", + "timepoint": 1964, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1964, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,syringe,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,needle,inserting; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1788_0", + "timepoint": 1788, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1788, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,syringe,holding; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_394_0", + "timepoint": 394, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 394, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound gel,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,instrument table,looking; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_526_0", + "timepoint": 526, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 526, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,assistant,closeto; patient,operating table,lyingon; assistant,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_1339_0", + "timepoint": 1339, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1339, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; needle,patient,touching; assistant,needle,removing; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1708_0", + "timepoint": 1708, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1708, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; assistant,operating table,closeto; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1579_0", + "timepoint": 1579, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1579, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_874_0", + "timepoint": 874, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 874, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1525_0", + "timepoint": 1525, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1525, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_194_0", + "timepoint": 194, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 194, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,patient,cutting; anesthetist,health monitor,looking; anesthetist,operating table,closeto; assistant,operating table,closeto; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; scalpel,patient,touching; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_3_6_616_0", + "timepoint": 616, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 616, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,curette,holding; curette,patient,touching; head surgeon,herbal disk,removing; head surgeon,scalpel,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; scalpel,patient,touching; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_1016_0", + "timepoint": 1016, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1016, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; patient,operating table,lyingon; assistant,patient,disinfection; assistant,operating table,closeto; head surgeon,operating table,closeto; cotton,patient,touching; assistant,antiseptic,holding; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2804_0", + "timepoint": 2804, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2804, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,syringe,injecting; syringe,needle,touching; needle,patient,touching; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1170_0", + "timepoint": 1170, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1170, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound probe,looking; assistant,needle,inserting; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_726_0", + "timepoint": 726, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 726, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2554_0", + "timepoint": 2554, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2554, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,tissue paper,applying; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2035_0", + "timepoint": 2035, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2035, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1147_0", + "timepoint": 1147, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1147, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; needle,patient,touching; assistant,ultrasound probe,looking; assistant,needle,inserting; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2295_0", + "timepoint": 2295, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2295, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,holding; assistant,instrument table,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2425_0", + "timepoint": 2425, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2425, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,needle,removing; head surgeon,ultrasound machine,looking; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_461_0", + "timepoint": 461, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 461, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1466_0", + "timepoint": 1466, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1466, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,microscope controller,manipulating; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_679_0", + "timepoint": 679, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 679, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_723_0", + "timepoint": 723, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 723, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,patient,manipulating; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1660_0", + "timepoint": 1660, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1660, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,looking; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_917_0", + "timepoint": 917, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 917, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,body marker,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2385_0", + "timepoint": 2385, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2385, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; needle,patient,touching; assistant,needle,removing; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1889_0", + "timepoint": 1889, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1889, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,syringe,injecting; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1222_0", + "timepoint": 1222, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1222, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_813_0", + "timepoint": 813, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 813, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_791_0", + "timepoint": 791, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 791, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; head surgeon,operating table,closeto; assistant,antiseptic,holding; forceps,cotton,touching; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1787_0", + "timepoint": 1787, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1787, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,needle,inserting; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,tissue mark,looking; head surgeon,ultrasound probe,controlling; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1000_0", + "timepoint": 1000, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1000, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,microscope,manipulating; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "MISS_3_6_928_0", + "timepoint": 928, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 928, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; assistant,patient,aspirating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1585_0", + "timepoint": 1585, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1585, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_201_0", + "timepoint": 201, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 201, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2773_0", + "timepoint": 2773, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2773, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; assistant,needle,removing; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_694_0", + "timepoint": 694, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 694, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; patient,operating table,lyingon; assistant,instrument table,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,curette,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_241_0", + "timepoint": 241, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 241, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_118_0", + "timepoint": 118, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 118, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1739_0", + "timepoint": 1739, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1739, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,tissue mark,looking; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1895_0", + "timepoint": 1895, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1895, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,syringe,injecting; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; syringe,needle,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_943_0", + "timepoint": 943, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 943, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_145_0", + "timepoint": 145, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 145, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto; circulator,scissors,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2000_0", + "timepoint": 2000, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2000, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_8_0", + "timepoint": 8, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 8, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_369_0", + "timepoint": 369, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 369, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; forceps,patient,touching; head surgeon,curette,holding; head surgeon,operating table,closeto; head surgeon,forceps,holding; head surgeon,herbal disk,removing; curette,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1228_0", + "timepoint": 1228, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1228, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_700_0", + "timepoint": 700, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 700, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,instrument table,looking; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; assistant,curette,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1522_0", + "timepoint": 1522, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1522, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,inserting; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1759_0", + "timepoint": 1759, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1759, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; assistant,syringe,holding; assistant,instrument table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_268_0", + "timepoint": 268, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 268, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; forceps,patient,touching; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,curette,holding; anesthetist,health monitor,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2510_0", + "timepoint": 2510, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2510, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2238_0", + "timepoint": 2238, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2238, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; cotton,patient,touching; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,patient,disinfection; forceps,cotton,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_118_0", + "timepoint": 118, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 118, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1060_0", + "timepoint": 1060, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1060, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; anesthetist,operating table,closeto; anesthetist,microscope,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2354_0", + "timepoint": 2354, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2354, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,syringe,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2590_0", + "timepoint": 2590, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2590, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; assistant,syringe,injecting; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; syringe,needle,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1624_0", + "timepoint": 1624, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1624, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,inserting; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2409_0", + "timepoint": 2409, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2409, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,ultrasound machine,looking; needle,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,needle,removing " + } + ] + }, + { + "id": "Ultrasound_3_4_517_0", + "timepoint": 517, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 517, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_60_0", + "timepoint": 60, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 60, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "MISS_1_1_433_0", + "timepoint": 433, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 433, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; anesthetist,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "MISS_1_1_1537_0", + "timepoint": 1537, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1537, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,patient,anaesthetising; head surgeon,microscope controller,manipulating " + } + ] + }, + { + "id": "MISS_3_6_1205_0", + "timepoint": 1205, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1205, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1903_0", + "timepoint": 1903, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1903, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,needle,holding; syringe,needle,touching; assistant,syringe,injecting; needle,patient,touching; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2256_0", + "timepoint": 2256, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2256, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2923_0", + "timepoint": 2923, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2923, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_529_0", + "timepoint": 529, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 529, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,assistant,closeto; assistant,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_2968_0", + "timepoint": 2968, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2968, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,scanning; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; needle,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,removing; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_119_0", + "timepoint": 119, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 119, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,syringe,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2323_0", + "timepoint": 2323, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2323, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_104_0", + "timepoint": 104, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 104, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,patient,manipulating; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_821_0", + "timepoint": 821, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 821, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; circulator,microscope,closeto; anesthetist,operating table,closeto; circulator,microscope,manipulating " + } + ] + }, + { + "id": "MISS_2_2_317_0", + "timepoint": 317, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 317, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; cotton,patient,touching; forceps,cotton,touching; anesthetist,health monitor,looking; head surgeon,operating table,closeto; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; assistant,patient,aspirating; assistant,forceps,holding; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1879_0", + "timepoint": 1879, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1879, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; assistant,syringe,injecting; head surgeon,needle,holding; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1852_0", + "timepoint": 1852, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1852, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon; assistant,syringe,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2815_0", + "timepoint": 2815, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2815, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; head surgeon,ultrasound machine,looking; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_349_0", + "timepoint": 349, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 349, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,manipulating; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_1344_0", + "timepoint": 1344, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1344, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1509_0", + "timepoint": 1509, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1509, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_471_0", + "timepoint": 471, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 471, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2484_0", + "timepoint": 2484, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2484, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_790_0", + "timepoint": 790, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 790, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; circulator,microscope,closeto; patient,operating table,lyingon; circulator,microscope,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_660_0", + "timepoint": 660, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 660, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_228_0", + "timepoint": 228, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 228, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1304_0", + "timepoint": 1304, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1304, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,antiseptic,holding; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_595_0", + "timepoint": 595, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 595, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1077_0", + "timepoint": 1077, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1077, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; patient,operating table,lyingon; assistant,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; anesthetist,operating table,closeto; anesthetist,health monitor,looking; cotton,patient,touching; head surgeon,operating table,closeto; assistant,patient,aspirating; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_629_0", + "timepoint": 629, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 629, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1309_0", + "timepoint": 1309, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1309, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,patient,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_512_0", + "timepoint": 512, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 512, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; curette,patient,touching; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; head surgeon,curette,holding; assistant,operating table,closeto; scalpel,patient,touching; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "Ultrasound_4_5_2465_0", + "timepoint": 2465, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2465, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound machine,looking; assistant,instrument table,closeto; assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3064_0", + "timepoint": 3064, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3064, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3279_0", + "timepoint": 3279, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3279, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,tissue paper,applying; tissue paper,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1872_0", + "timepoint": 1872, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1872, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2614_0", + "timepoint": 2614, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2614, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,tissue paper,applying; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2713_0", + "timepoint": 2713, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2713, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,bin,closeto; assistant,tissue paper,dropping; patient,operating table,lyingon; assistant,bin,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2486_0", + "timepoint": 2486, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2486, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; syringe,needle,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,syringe,injecting; assistant,needle,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1293_0", + "timepoint": 1293, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1293, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_2462_0", + "timepoint": 2462, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2462, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,looking; patient,operating table,lyingon; assistant,instrument table,closeto; assistant,needle,holding; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "MISS_3_6_442_0", + "timepoint": 442, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 442, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,curette,holding; assistant,scalpel,holding; head surgeon,forceps,holding; curette,patient,touching; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "Ultrasound_3_4_3052_0", + "timepoint": 3052, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3052, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_3_4_396_0", + "timepoint": 396, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 396, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1254_0", + "timepoint": 1254, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1254, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2079_0", + "timepoint": 2079, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2079, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2383_0", + "timepoint": 2383, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2383, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,removing; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_690_0", + "timepoint": 690, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 690, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_196_0", + "timepoint": 196, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 196, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_991_0", + "timepoint": 991, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 991, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1551_0", + "timepoint": 1551, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1551, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_3_4_887_0", + "timepoint": 887, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 887, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound probe,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; body marker,patient,touching; assistant,body marker,applying; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1219_0", + "timepoint": 1219, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1219, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,inserting; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_425_0", + "timepoint": 425, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 425, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; head surgeon,herbal disk,removing; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,forceps,holding; head surgeon,operating table,closeto; head surgeon,curette,holding; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; assistant,instrument table,looking; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_234_0", + "timepoint": 234, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 234, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_2063_0", + "timepoint": 2063, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2063, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; patient,operating table,lyingon; needle,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_604_0", + "timepoint": 604, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 604, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_2809_0", + "timepoint": 2809, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2809, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; assistant,needle,holding; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_1_1_708_0", + "timepoint": 708, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 708, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_734_0", + "timepoint": 734, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 734, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; assistant,microscope screen,looking; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_446_0", + "timepoint": 446, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 446, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_928_0", + "timepoint": 928, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 928, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,body marker,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_858_0", + "timepoint": 858, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 858, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2158_0", + "timepoint": 2158, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2158, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,disinfection; patient,operating table,lyingon; assistant,operating table,closeto; forceps,cotton,touching; cotton,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_684_0", + "timepoint": 684, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 684, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,looking; assistant,instrument table,closeto; head surgeon,microscope eye,looking; assistant,curette,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1065_0", + "timepoint": 1065, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1065, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2628_0", + "timepoint": 2628, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2628, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; syringe,needle,touching; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_64_0", + "timepoint": 64, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 64, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1787_0", + "timepoint": 1787, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1787, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1842_0", + "timepoint": 1842, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1842, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,antiseptic,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_2608_0", + "timepoint": 2608, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2608, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,tissue paper,applying; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2111_0", + "timepoint": 2111, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2111, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; syringe,needle,touching; head surgeon,needle,holding; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_703_0", + "timepoint": 703, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 703, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1642_0", + "timepoint": 1642, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1642, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,tissue mark,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2844_0", + "timepoint": 2844, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2844, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_988_0", + "timepoint": 988, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 988, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_998_0", + "timepoint": 998, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 998, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,body marker,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_675_0", + "timepoint": 675, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 675, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; head surgeon,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1255_0", + "timepoint": 1255, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1255, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,patient,manipulating; anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1202_0", + "timepoint": 1202, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1202, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating room,entering; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1041_0", + "timepoint": 1041, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1041, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1492_0", + "timepoint": 1492, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1492, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_2213_0", + "timepoint": 2213, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2213, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; cotton,patient,touching; assistant,patient,disinfection; assistant,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_1109_0", + "timepoint": 1109, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1109, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1960_0", + "timepoint": 1960, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1960, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,instrument table,closeto; head surgeon,needle,holding; assistant,syringe,holding; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1014_0", + "timepoint": 1014, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1014, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,microscope,manipulating; anesthetist,operating table,closeto; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_2916_0", + "timepoint": 2916, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2916, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,holding; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1463_0", + "timepoint": 1463, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1463, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_126_0", + "timepoint": 126, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 126, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; anesthetist,operating table,closeto; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,patient,cutting; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1927_0", + "timepoint": 1927, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1927, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,needle,holding; assistant,instrument table,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1323_0", + "timepoint": 1323, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1323, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; assistant,operating table,closeto; assistant,body marker,applying; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; body marker,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_49_0", + "timepoint": 49, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 49, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,curette,holding; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "MISS_1_1_2357_0", + "timepoint": 2357, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2357, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1304_0", + "timepoint": 1304, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1304, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,needle,inserting; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1540_0", + "timepoint": 1540, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1540, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_1139_0", + "timepoint": 1139, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1139, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,health monitor,looking; anesthetist,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1188_0", + "timepoint": 1188, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1188, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2723_0", + "timepoint": 2723, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2723, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,bin,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_364_0", + "timepoint": 364, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 364, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; head surgeon,operating table,closeto; curette,patient,touching; head surgeon,curette,holding; assistant,operating table,closeto; head surgeon,forceps,holding; patient,operating table,lyingon; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; forceps,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_569_0", + "timepoint": 569, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 569, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1238_0", + "timepoint": 1238, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1238, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; patient,operating table,lyingon; assistant,body marker,applying; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2447_0", + "timepoint": 2447, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2447, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; assistant,syringe,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1570_0", + "timepoint": 1570, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1570, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; assistant,ultrasound probe,looking; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_2_2_166_0", + "timepoint": 166, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 166, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_942_0", + "timepoint": 942, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 942, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; patient,operating table,lyingon; circulator,microscope,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_233_0", + "timepoint": 233, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 233, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,scissors,looking; circulator,instruments,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1025_0", + "timepoint": 1025, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1025, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1869_0", + "timepoint": 1869, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1869, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1075_0", + "timepoint": 1075, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1075, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,aspirating; head surgeon,microscope eye,looking; patient,operating table,lyingon; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; cotton,patient,touching; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2123_0", + "timepoint": 2123, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2123, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; assistant,syringe,injecting; ultrasound probe,patient,touching; syringe,needle,touching; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3163_0", + "timepoint": 3163, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3163, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; tissue paper,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_1718_0", + "timepoint": 1718, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1718, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,operating table,closeto; assistant,needle,inserting " + } + ] + }, + { + "id": "MISS_1_1_882_0", + "timepoint": 882, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 882, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,microscope,manipulating; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1121_0", + "timepoint": 1121, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1121, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_293_0", + "timepoint": 293, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 293, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2211_0", + "timepoint": 2211, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2211, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,syringe,injecting; assistant,needle,holding; head surgeon,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1815_0", + "timepoint": 1815, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1815, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,holding; assistant,needle,holding; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_502_0", + "timepoint": 502, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 502, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_699_0", + "timepoint": 699, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 699, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1095_0", + "timepoint": 1095, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1095, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,scissors,manipulating; assistant,patient,disinfection; assistant,operating table,closeto; scissors,cotton,touching; assistant,antiseptic,holding; cotton,patient,touching; patient,operating table,lyingon; circulator,ultrasound machine,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2387_0", + "timepoint": 2387, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2387, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,needle,removing; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1525_0", + "timepoint": 1525, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1525, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_645_0", + "timepoint": 645, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 645, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2060_0", + "timepoint": 2060, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2060, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,needle,holding; needle,patient,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_5_6_3075_0", + "timepoint": 3075, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3075, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_66_0", + "timepoint": 66, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 66, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_1_1_664_0", + "timepoint": 664, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 664, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1213_0", + "timepoint": 1213, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1213, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_428_0", + "timepoint": 428, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 428, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_463_0", + "timepoint": 463, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 463, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1846_0", + "timepoint": 1846, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1846, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,antiseptic,applying; assistant,patient,disinfection; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_842_0", + "timepoint": 842, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 842, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_3247_0", + "timepoint": 3247, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3247, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,applying; patient,operating table,lyingon; tissue paper,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1505_0", + "timepoint": 1505, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1505, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2372_0", + "timepoint": 2372, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2372, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,needle,holding; assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_626_0", + "timepoint": 626, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 626, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2205_0", + "timepoint": 2205, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2205, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,syringe,injecting; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound screen,looking; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_472_0", + "timepoint": 472, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 472, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_525_0", + "timepoint": 525, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 525, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_385_0", + "timepoint": 385, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 385, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "MISS_1_1_806_0", + "timepoint": 806, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 806, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,closeto; anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_370_0", + "timepoint": 370, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 370, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,closeto; circulator,operating table,looking; circulator,patient,closeto " + } + ] + }, + { + "id": "MISS_1_1_2494_0", + "timepoint": 2494, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2494, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1486_0", + "timepoint": 1486, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1486, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2403_0", + "timepoint": 2403, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2403, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; syringe,needle,touching; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,needle,holding; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_428_0", + "timepoint": 428, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 428, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; head surgeon,operating table,closeto; assistant,scalpel,holding; anesthetist,operating table,closeto; assistant,instrument table,closeto; head surgeon,forceps,holding; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; curette,patient,touching; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_536_0", + "timepoint": 536, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 536, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating room,entering; patient,operating table,lyingon; circulator,assistant,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2108_0", + "timepoint": 2108, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2108, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; syringe,needle,touching; needle,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; assistant,syringe,injecting; assistant,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1615_0", + "timepoint": 1615, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1615, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,needle,inserting; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,tissue mark,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_486_0", + "timepoint": 486, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 486, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,scalpel,holding; assistant,instrument table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3274_0", + "timepoint": 3274, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3274, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,tissue paper,applying; patient,operating table,lyingon; tissue paper,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_513_0", + "timepoint": 513, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 513, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; anesthetist,operating table,closeto; circulator,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope controller,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2281_0", + "timepoint": 2281, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2281, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,needle,holding; syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2722_0", + "timepoint": 2722, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2722, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,bin,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1326_0", + "timepoint": 1326, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1326, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2244_0", + "timepoint": 2244, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2244, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,ultrasound screen,looking; assistant,operating table,closeto; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_5_6_1445_0", + "timepoint": 1445, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1445, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1111_0", + "timepoint": 1111, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1111, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2856_0", + "timepoint": 2856, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2856, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; assistant,syringe,removing " + } + ] + }, + { + "id": "Ultrasound_4_5_2384_0", + "timepoint": 2384, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2384, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_914_0", + "timepoint": 914, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 914, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1453_0", + "timepoint": 1453, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1453, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_308_0", + "timepoint": 308, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 308, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; head surgeon,forceps,holding; head surgeon,curette,holding; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; curette,patient,touching; head surgeon,microscope eye,looking; patient,operating table,lyingon; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; forceps,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2421_0", + "timepoint": 2421, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2421, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound machine,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,needle,removing; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_703_0", + "timepoint": 703, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 703, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,looking; assistant,instrument table,closeto; circulator,ultrasound machine,closeto; circulator,ultrasound machine,manipulating; patient,operating table,lyingon; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "MISS_2_2_223_0", + "timepoint": 223, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 223, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,operating table,closeto; anesthetist,health monitor,looking; assistant,operating table,closeto; anesthetist,operating table,closeto; assistant,forceps,holding; head surgeon,patient,cutting; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; scalpel,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2783_0", + "timepoint": 2783, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2783, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2923_0", + "timepoint": 2923, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2923, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,syringe,holding; assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2286_0", + "timepoint": 2286, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2286, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,needle,holding; needle,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound probe,controlling; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1983_0", + "timepoint": 1983, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1983, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,syringe,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1849_0", + "timepoint": 1849, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1849, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_1746_0", + "timepoint": 1746, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1746, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,antiseptic,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2253_0", + "timepoint": 2253, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2253, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,forceps,holding; assistant,operating table,closeto; assistant,patient,disinfection; forceps,cotton,touching; cotton,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1902_0", + "timepoint": 1902, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1902, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1005_0", + "timepoint": 1005, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1005, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; head surgeon,operating table,closeto; cotton,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_4_5_270_0", + "timepoint": 270, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 270, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound gel,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1865_0", + "timepoint": 1865, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1865, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_646_0", + "timepoint": 646, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 646, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_268_0", + "timepoint": 268, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 268, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1884_0", + "timepoint": 1884, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1884, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; needle,patient,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_1855_0", + "timepoint": 1855, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1855, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,antiseptic,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_4_5_147_0", + "timepoint": 147, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 147, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_1_1_1848_0", + "timepoint": 1848, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1848, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; anesthetist,operating table,closeto; assistant,antiseptic,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_958_0", + "timepoint": 958, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 958, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_258_0", + "timepoint": 258, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 258, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_853_0", + "timepoint": 853, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 853, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_204_0", + "timepoint": 204, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 204, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; head surgeon,herbal disk,removing; head surgeon,forceps,holding; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,curette,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1247_0", + "timepoint": 1247, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1247, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2328_0", + "timepoint": 2328, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2328, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2449_0", + "timepoint": 2449, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2449, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,needle,holding; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_673_0", + "timepoint": 673, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 673, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "MISS_1_1_1232_0", + "timepoint": 1232, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1232, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; anesthetist,patient,manipulating; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2391_0", + "timepoint": 2391, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2391, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,syringe,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2617_0", + "timepoint": 2617, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2617, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1875_0", + "timepoint": 1875, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1875, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,antiseptic,applying; anesthetist,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_4_5_904_0", + "timepoint": 904, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 904, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2054_0", + "timepoint": 2054, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2054, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,needle,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_663_0", + "timepoint": 663, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 663, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_329_0", + "timepoint": 329, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 329, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; forceps,patient,touching; head surgeon,microscope eye,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,curette,holding; head surgeon,herbal disk,removing; curette,patient,touching; anesthetist,health monitor,looking; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2275_0", + "timepoint": 2275, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2275, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,needle,holding; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,needle,holding; needle,patient,touching; syringe,needle,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_282_0", + "timepoint": 282, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 282, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2133_0", + "timepoint": 2133, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2133, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,syringe,injecting; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,ultrasound screen,looking; assistant,operating table,closeto; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2955_0", + "timepoint": 2955, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2955, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,patient,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,syringe,holding; assistant,needle,removing; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1208_0", + "timepoint": 1208, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1208, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto; assistant,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2905_0", + "timepoint": 2905, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2905, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "MISS_2_2_622_0", + "timepoint": 622, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 622, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,microscope controller,controlling; assistant,operating table,closeto; patient,operating table,lyingon; assistant,microscope screen,looking; anesthetist,health monitor,looking; head surgeon,microscope screen,looking; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_697_0", + "timepoint": 697, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 697, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_576_0", + "timepoint": 576, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 576, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "MISS_3_6_257_0", + "timepoint": 257, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 257, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,curette,holding; head surgeon,forceps,holding; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; patient,operating table,lyingon; curette,patient,touching; head surgeon,herbal disk,removing; anesthetist,health monitor,looking; forceps,patient,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_460_0", + "timepoint": 460, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 460, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_93_0", + "timepoint": 93, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 93, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_881_0", + "timepoint": 881, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 881, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_983_0", + "timepoint": 983, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 983, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2986_0", + "timepoint": 2986, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2986, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_441_0", + "timepoint": 441, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 441, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_560_0", + "timepoint": 560, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 560, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,health monitor,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,microscope controller,controlling; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1895_0", + "timepoint": 1895, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1895, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,antiseptic,applying; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_3_4_1891_0", + "timepoint": 1891, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1891, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,needle,holding; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,holding; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_640_0", + "timepoint": 640, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 640, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_1269_0", + "timepoint": 1269, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1269, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,applying; ultrasound probe,patient,touching; body marker,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_450_0", + "timepoint": 450, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 450, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1720_0", + "timepoint": 1720, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1720, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,syringe,holding; ultrasound probe,patient,touching; assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2610_0", + "timepoint": 2610, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2610, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,needle,holding; syringe,needle,touching; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_485_0", + "timepoint": 485, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 485, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,scalpel,holding; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2706_0", + "timepoint": 2706, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2706, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; needle,patient,touching; syringe,needle,touching; assistant,operating table,closeto; assistant,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1015_0", + "timepoint": 1015, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1015, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,body marker,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1403_0", + "timepoint": 1403, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1403, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1708_0", + "timepoint": 1708, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1708, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,instrument table,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_568_0", + "timepoint": 568, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 568, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_521_0", + "timepoint": 521, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 521, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,assistant,closeto; patient,operating table,lyingon; assistant,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_2_2_875_0", + "timepoint": 875, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 875, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,antiseptic,applying; assistant,patient,disinfection; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_596_0", + "timepoint": 596, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 596, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_138_0", + "timepoint": 138, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 138, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,curette,holding; head surgeon,herbal disk,removing; curette,patient,touching; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_936_0", + "timepoint": 936, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 936, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_617_0", + "timepoint": 617, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 617, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_193_0", + "timepoint": 193, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 193, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2513_0", + "timepoint": 2513, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2513, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,needle,holding; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_264_0", + "timepoint": 264, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 264, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_647_0", + "timepoint": 647, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 647, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,operating table,closeto; assistant,curette,holding; assistant,scalpel,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1005_0", + "timepoint": 1005, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1005, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,patient,aspirating; head surgeon,operating table,closeto; cotton,patient,touching; anesthetist,health monitor,looking; forceps,cotton,touching; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1060_0", + "timepoint": 1060, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1060, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_759_0", + "timepoint": 759, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 759, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1316_0", + "timepoint": 1316, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1316, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2294_0", + "timepoint": 2294, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2294, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning; syringe,needle,touching; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_412_0", + "timepoint": 412, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 412, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,instrument table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2206_0", + "timepoint": 2206, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2206, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; assistant,operating table,closeto; patient,operating table,lyingon; syringe,needle,touching; assistant,needle,holding; needle,patient,touching; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_335_0", + "timepoint": 335, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 335, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound gel,applying; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1534_0", + "timepoint": 1534, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1534, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1767_0", + "timepoint": 1767, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1767, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,needle,holding; needle,patient,touching; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_50_0", + "timepoint": 50, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 50, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; assistant,instrument table,closeto; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3308_0", + "timepoint": 3308, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3308, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,tissue paper,dropping; assistant,bin,looking; assistant,bin,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_804_0", + "timepoint": 804, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 804, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1833_0", + "timepoint": 1833, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1833, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_774_0", + "timepoint": 774, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 774, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1189_0", + "timepoint": 1189, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1189, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1013_0", + "timepoint": 1013, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1013, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2194_0", + "timepoint": 2194, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2194, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_2_2_4_0", + "timepoint": 4, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 4, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3235_0", + "timepoint": 3235, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3235, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,tissue paper,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_755_0", + "timepoint": 755, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 755, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_355_0", + "timepoint": 355, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 355, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,applying; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_438_0", + "timepoint": 438, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 438, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1308_0", + "timepoint": 1308, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1308, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2020_0", + "timepoint": 2020, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2020, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,syringe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_202_0", + "timepoint": 202, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 202, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_440_0", + "timepoint": 440, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 440, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1853_0", + "timepoint": 1853, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1853, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,syringe,holding; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_663_0", + "timepoint": 663, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 663, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,operating table,manipulating " + } + ] + }, + { + "id": "MISS_1_1_1868_0", + "timepoint": 1868, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1868, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,antiseptic,applying; assistant,operating table,closeto; assistant,patient,disinfection; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_763_0", + "timepoint": 763, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 763, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_458_0", + "timepoint": 458, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 458, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,scalpel,holding; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; assistant,instrument table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1230_0", + "timepoint": 1230, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1230, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2937_0", + "timepoint": 2937, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2937, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,tissue paper,applying; tissue paper,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1970_0", + "timepoint": 1970, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1970, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,needle,holding; assistant,syringe,holding; assistant,instrument table,closeto; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1536_0", + "timepoint": 1536, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1536, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,inserting; assistant,operating table,closeto; assistant,ultrasound probe,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_514_0", + "timepoint": 514, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 514, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3253_0", + "timepoint": 3253, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3253, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; tissue paper,patient,touching; patient,operating table,lyingon; assistant,tissue paper,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2568_0", + "timepoint": 2568, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2568, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; syringe,needle,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,syringe,injecting; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_609_0", + "timepoint": 609, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 609, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1176_0", + "timepoint": 1176, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1176, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1304_0", + "timepoint": 1304, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1304, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; body marker,patient,touching; head surgeon,ultrasound probe,controlling; assistant,ultrasound probe,looking; ultrasound probe,patient,touching; assistant,body marker,applying; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_702_0", + "timepoint": 702, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 702, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2075_0", + "timepoint": 2075, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2075, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "MISS_2_2_612_0", + "timepoint": 612, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 612, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,microscope screen,looking; head surgeon,microscope screen,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,health monitor,looking; head surgeon,microscope controller,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_736_0", + "timepoint": 736, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 736, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1022_0", + "timepoint": 1022, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1022, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,body marker,holding; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1750_0", + "timepoint": 1750, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1750, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,instrument table,looking; needle,patient,touching; ultrasound probe,patient,touching; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_1182_0", + "timepoint": 1182, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1182, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2502_0", + "timepoint": 2502, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2502, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1110_0", + "timepoint": 1110, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1110, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,body marker,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_2415_0", + "timepoint": 2415, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2415, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,ultrasound machine,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,needle,removing; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_543_0", + "timepoint": 543, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 543, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,herbal disk,removing; assistant,operating table,closeto; head surgeon,curette,holding; head surgeon,operating table,closeto; head surgeon,scalpel,holding; patient,operating table,lyingon; curette,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; scalpel,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_43_0", + "timepoint": 43, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 43, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3327_0", + "timepoint": 3327, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3327, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1421_0", + "timepoint": 1421, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1421, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_823_0", + "timepoint": 823, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 823, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,microscope,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "MISS_1_1_925_0", + "timepoint": 925, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 925, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; circulator,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2333_0", + "timepoint": 2333, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2333, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,syringe,holding; ultrasound probe,patient,touching; needle,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_85_0", + "timepoint": 85, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 85, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling; circulator,curette,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2576_0", + "timepoint": 2576, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2576, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,needle,holding; head surgeon,operating table,closeto; syringe,needle,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1588_0", + "timepoint": 1588, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1588, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,needle,holding; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_929_0", + "timepoint": 929, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 929, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2105_0", + "timepoint": 2105, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2105, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_2_2_395_0", + "timepoint": 395, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 395, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,aspirating; assistant,forceps,holding; cotton,patient,touching; assistant,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_708_0", + "timepoint": 708, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 708, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_3116_0", + "timepoint": 3116, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3116, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_230_0", + "timepoint": 230, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 230, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; curette,patient,touching; head surgeon,curette,holding; head surgeon,forceps,holding; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_397_0", + "timepoint": 397, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 397, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_948_0", + "timepoint": 948, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 948, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_589_0", + "timepoint": 589, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 589, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,health monitor,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,microscope controller,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1317_0", + "timepoint": 1317, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1317, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; body marker,patient,touching; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2066_0", + "timepoint": 2066, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2066, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,needle,holding; patient,operating table,lyingon; syringe,needle,touching; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,syringe,injecting; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1570_0", + "timepoint": 1570, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1570, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,needle,holding; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_717_0", + "timepoint": 717, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 717, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; anesthetist,operating table,closeto; assistant,microscope screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_877_0", + "timepoint": 877, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 877, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound probe,looking; patient,operating table,lyingon; assistant,body marker,applying; body marker,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_217_0", + "timepoint": 217, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 217, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_2173_0", + "timepoint": 2173, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2173, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,holding; ultrasound probe,patient,touching; needle,patient,touching; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_287_0", + "timepoint": 287, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 287, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_460_0", + "timepoint": 460, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 460, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1648_0", + "timepoint": 1648, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1648, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,instrument table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,instrument table,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_238_0", + "timepoint": 238, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 238, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; assistant,operating table,closeto; head surgeon,scalpel,holding; anesthetist,operating table,closeto; head surgeon,patient,cutting; forceps,cotton,touching; head surgeon,operating table,closeto; scalpel,patient,touching; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_370_0", + "timepoint": 370, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 370, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "MISS_2_2_121_0", + "timepoint": 121, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 121, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto; head surgeon,scalpel,holding; assistant,instrument table,closeto; head surgeon,patient,cutting; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2651_0", + "timepoint": 2651, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2651, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; syringe,needle,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,needle,holding; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_634_0", + "timepoint": 634, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 634, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1151_0", + "timepoint": 1151, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1151, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_311_0", + "timepoint": 311, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 311, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,herbal disk,removing; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,forceps,holding; curette,patient,touching; patient,operating table,lyingon; forceps,patient,touching; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1762_0", + "timepoint": 1762, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1762, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,syringe,holding; head surgeon,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2087_0", + "timepoint": 2087, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2087, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,needle,holding; assistant,syringe,injecting; syringe,needle,touching; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2186_0", + "timepoint": 2186, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2186, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,holding; syringe,needle,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_71_0", + "timepoint": 71, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 71, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,curette,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1264_0", + "timepoint": 1264, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1264, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2607_0", + "timepoint": 2607, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2607, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,tissue paper,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2627_0", + "timepoint": 2627, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2627, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1198_0", + "timepoint": 1198, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1198, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2031_0", + "timepoint": 2031, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2031, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,syringe,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_196_0", + "timepoint": 196, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 196, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,ultrasound gel,applying; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_764_0", + "timepoint": 764, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 764, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; forceps,cotton,touching; patient,operating table,lyingon; assistant,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2208_0", + "timepoint": 2208, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2208, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,operating table,closeto; cotton,patient,touching; patient,operating table,lyingon; assistant,patient,disinfection; assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3025_0", + "timepoint": 3025, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3025, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; tissue paper,patient,touching; head surgeon,tissue paper,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1407_0", + "timepoint": 1407, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1407, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; assistant,body marker,holding; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_809_0", + "timepoint": 809, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 809, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,body marker,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_412_0", + "timepoint": 412, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 412, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,instrument table,looking; assistant,operating table,closeto; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_678_0", + "timepoint": 678, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 678, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_328_0", + "timepoint": 328, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 328, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; circulator,operating table,looking; patient,operating room,entering " + } + ] + }, + { + "id": "MISS_1_1_2185_0", + "timepoint": 2185, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2185, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; cotton,patient,touching; head surgeon,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,forceps,holding; assistant,patient,disinfection " + } + ] + }, + { + "id": "MISS_3_6_947_0", + "timepoint": 947, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 947, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; cotton,patient,touching; forceps,cotton,touching; head surgeon,microscope eye,looking; assistant,patient,aspirating; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_713_0", + "timepoint": 713, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 713, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2608_0", + "timepoint": 2608, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2608, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,holding; syringe,needle,touching; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_3_6_399_0", + "timepoint": 399, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 399, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,curette,holding; head surgeon,forceps,holding; curette,patient,touching; assistant,forceps,holding; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2095_0", + "timepoint": 2095, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2095, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; syringe,needle,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1276_0", + "timepoint": 1276, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1276, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,body marker,applying; body marker,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2361_0", + "timepoint": 2361, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2361, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2163_0", + "timepoint": 2163, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2163, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2883_0", + "timepoint": 2883, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2883, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,removing; head surgeon,operating table,closeto; syringe,needle,touching; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_751_0", + "timepoint": 751, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 751, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2324_0", + "timepoint": 2324, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2324, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,needle,holding; needle,patient,touching; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_2_2_70_0", + "timepoint": 70, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 70, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,scalpel,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,patient,manipulating; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_738_0", + "timepoint": 738, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 738, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,instrument table,closeto; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3151_0", + "timepoint": 3151, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3151, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,applying; assistant,operating table,closeto; tissue paper,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1195_0", + "timepoint": 1195, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1195, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2052_0", + "timepoint": 2052, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2052, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto; assistant,patient,disinfection; head surgeon,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_2_2_38_0", + "timepoint": 38, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 38, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1990_0", + "timepoint": 1990, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1990, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; assistant,syringe,injecting; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1980_0", + "timepoint": 1980, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1980, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,syringe,injecting; syringe,needle,touching; assistant,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1613_0", + "timepoint": 1613, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1613, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,inserting; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_319_0", + "timepoint": 319, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 319, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,applying " + } + ] + }, + { + "id": "MISS_1_1_1267_0", + "timepoint": 1267, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1267, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,patient,manipulating; anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1741_0", + "timepoint": 1741, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1741, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,syringe,holding; assistant,needle,holding; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2118_0", + "timepoint": 2118, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2118, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; anesthetist,operating table,closeto; cotton,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,disinfection; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_344_0", + "timepoint": 344, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 344, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,applying; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2103_0", + "timepoint": 2103, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2103, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; assistant,operating table,closeto; needle,patient,touching; assistant,syringe,injecting; syringe,needle,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_5_0", + "timepoint": 5, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 5, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_70_0", + "timepoint": 70, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 70, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound machine,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_983_0", + "timepoint": 983, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 983, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; forceps,cotton,touching; assistant,patient,disinfection; assistant,forceps,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_621_0", + "timepoint": 621, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 621, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1921_0", + "timepoint": 1921, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1921, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,antiseptic,applying; assistant,patient,disinfection; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_641_0", + "timepoint": 641, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 641, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,curette,holding; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; assistant,scalpel,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2478_0", + "timepoint": 2478, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2478, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,syringe,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2152_0", + "timepoint": 2152, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2152, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; patient,operating table,lyingon; cotton,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,disinfection; forceps,cotton,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_478_0", + "timepoint": 478, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 478, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_81_0", + "timepoint": 81, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 81, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1707_0", + "timepoint": 1707, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1707, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1154_0", + "timepoint": 1154, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1154, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,ultrasound machine,manipulating; assistant,antiseptic,holding; assistant,forceps,holding; circulator,ultrasound machine,closeto; forceps,cotton,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_92_0", + "timepoint": 92, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 92, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling; circulator,body marker,looking " + } + ] + }, + { + "id": "MISS_3_6_106_0", + "timepoint": 106, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 106, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; assistant,operating table,closeto; head surgeon,curette,holding; anesthetist,operating table,closeto; curette,patient,touching; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1597_0", + "timepoint": 1597, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1597, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_754_0", + "timepoint": 754, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 754, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1520_0", + "timepoint": 1520, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1520, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2535_0", + "timepoint": 2535, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2535, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_562_0", + "timepoint": 562, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 562, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1861_0", + "timepoint": 1861, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1861, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,antiseptic,applying; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_242_0", + "timepoint": 242, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 242, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,instrument table,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2642_0", + "timepoint": 2642, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2642, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,holding; assistant,operating table,closeto; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; syringe,needle,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_1105_0", + "timepoint": 1105, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1105, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3211_0", + "timepoint": 3211, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3211, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; tissue paper,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; assistant,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_3_4_779_0", + "timepoint": 779, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 779, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,instrument table,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1177_0", + "timepoint": 1177, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1177, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_889_0", + "timepoint": 889, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 889, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; forceps,cotton,touching; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1385_0", + "timepoint": 1385, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1385, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1640_0", + "timepoint": 1640, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1640, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,needle,inserting; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,tissue mark,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_146_0", + "timepoint": 146, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 146, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; anesthetist,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; curette,patient,touching; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1503_0", + "timepoint": 1503, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1503, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,needle,inserting; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1634_0", + "timepoint": 1634, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1634, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1672_0", + "timepoint": 1672, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1672, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,instrument table,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_3_6_468_0", + "timepoint": 468, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 468, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; head surgeon,curette,holding; head surgeon,herbal disk,removing; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; curette,patient,touching; head surgeon,forceps,holding; assistant,operating table,closeto; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2482_0", + "timepoint": 2482, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2482, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; assistant,needle,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,needle,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; syringe,needle,touching; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1698_0", + "timepoint": 1698, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1698, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,inserting; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,tissue mark,looking; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2537_0", + "timepoint": 2537, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2537, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,syringe,injecting; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_192_0", + "timepoint": 192, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 192, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound gel,applying; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_1_1_660_0", + "timepoint": 660, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 660, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_3068_0", + "timepoint": 3068, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3068, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,tissue paper,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1792_0", + "timepoint": 1792, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1792, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,antiseptic,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2300_0", + "timepoint": 2300, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2300, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1471_0", + "timepoint": 1471, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1471, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1791_0", + "timepoint": 1791, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1791, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,syringe,holding; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2657_0", + "timepoint": 2657, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2657, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; assistant,operating table,closeto; syringe,needle,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; needle,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1009_0", + "timepoint": 1009, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1009, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,body marker,holding " + } + ] + }, + { + "id": "MISS_3_6_1035_0", + "timepoint": 1035, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1035, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,patient,aspirating; head surgeon,operating table,closeto; cotton,patient,touching; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,health monitor,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_422_0", + "timepoint": 422, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 422, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_816_0", + "timepoint": 816, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 816, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; circulator,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,microscope,closeto " + } + ] + }, + { + "id": "MISS_3_6_1150_0", + "timepoint": 1150, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1150, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_103_0", + "timepoint": 103, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 103, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,ultrasound gel,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2601_0", + "timepoint": 2601, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2601, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; syringe,needle,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,needle,holding; needle,patient,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_3045_0", + "timepoint": 3045, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3045, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,tissue paper,holding; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_3_6_670_0", + "timepoint": 670, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 670, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,looking; patient,operating table,lyingon; assistant,scalpel,holding; assistant,curette,holding; head surgeon,microscope eye,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_633_0", + "timepoint": 633, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 633, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1018_0", + "timepoint": 1018, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1018, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_2_2_799_0", + "timepoint": 799, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 799, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,microscope screen,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,health monitor,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_491_0", + "timepoint": 491, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 491, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2189_0", + "timepoint": 2189, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2189, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,needle,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1157_0", + "timepoint": 1157, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1157, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_101_0", + "timepoint": 101, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 101, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_405_0", + "timepoint": 405, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 405, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1268_0", + "timepoint": 1268, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1268, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; assistant,needle,inserting; assistant,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_135_0", + "timepoint": 135, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 135, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound machine,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2480_0", + "timepoint": 2480, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2480, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,patient,manipulating; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_367_0", + "timepoint": 367, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 367, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2025_0", + "timepoint": 2025, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2025, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_3238_0", + "timepoint": 3238, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3238, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; tissue paper,patient,touching; assistant,tissue paper,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_502_0", + "timepoint": 502, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 502, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "MISS_3_6_465_0", + "timepoint": 465, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 465, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,forceps,holding; curette,patient,touching; head surgeon,curette,holding; head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,herbal disk,removing; head surgeon,scalpel,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_303_0", + "timepoint": 303, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 303, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; forceps,patient,touching; anesthetist,health monitor,looking; head surgeon,forceps,holding; assistant,operating table,closeto; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2971_0", + "timepoint": 2971, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2971, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,scanning; head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; assistant,needle,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_683_0", + "timepoint": 683, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 683, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_3064_0", + "timepoint": 3064, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3064, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_138_0", + "timepoint": 138, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 138, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto; circulator,scalpel,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_104_0", + "timepoint": 104, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 104, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_419_0", + "timepoint": 419, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 419, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_872_0", + "timepoint": 872, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 872, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2348_0", + "timepoint": 2348, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2348, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching; assistant,needle,removing; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_106_0", + "timepoint": 106, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 106, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon; assistant,patient,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2264_0", + "timepoint": 2264, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2264, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound screen,looking; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_67_0", + "timepoint": 67, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 67, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,patient,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_374_0", + "timepoint": 374, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 374, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1671_0", + "timepoint": 1671, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1671, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1869_0", + "timepoint": 1869, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1869, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,antiseptic,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_5_6_2940_0", + "timepoint": 2940, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2940, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,needle,removing; head surgeon,ultrasound probe,controlling; head surgeon,patient,scanning; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1402_0", + "timepoint": 1402, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1402, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2315_0", + "timepoint": 2315, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2315, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; syringe,needle,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2146_0", + "timepoint": 2146, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2146, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,syringe,injecting; assistant,operating table,closeto; syringe,needle,touching; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1369_0", + "timepoint": 1369, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1369, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,needle,removing; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_127_0", + "timepoint": 127, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 127, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1895_0", + "timepoint": 1895, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1895, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; needle,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_603_0", + "timepoint": 603, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 603, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; scalpel,patient,touching; curette,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,herbal disk,removing; head surgeon,curette,holding; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1311_0", + "timepoint": 1311, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1311, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2417_0", + "timepoint": 2417, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2417, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,syringe,injecting; assistant,needle,holding; syringe,needle,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,needle,holding; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_350_0", + "timepoint": 350, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 350, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1797_0", + "timepoint": 1797, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1797, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_37_0", + "timepoint": 37, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 37, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2378_0", + "timepoint": 2378, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2378, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_148_0", + "timepoint": 148, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 148, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling; circulator,scissors,looking " + } + ] + }, + { + "id": "MISS_2_2_511_0", + "timepoint": 511, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 511, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; head surgeon,microscope controller,controlling; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; circulator,instrument table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1202_0", + "timepoint": 1202, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1202, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_167_0", + "timepoint": 167, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 167, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; scalpel,patient,touching; forceps,cotton,touching; assistant,instrument table,closeto; head surgeon,patient,cutting; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_628_0", + "timepoint": 628, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 628, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_655_0", + "timepoint": 655, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 655, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1255_0", + "timepoint": 1255, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1255, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding; forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1886_0", + "timepoint": 1886, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1886, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,needle,holding; head surgeon,needle,holding; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_969_0", + "timepoint": 969, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 969, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1159_0", + "timepoint": 1159, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1159, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,needle,inserting; needle,patient,touching; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,ultrasound probe,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_776_0", + "timepoint": 776, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 776, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_980_0", + "timepoint": 980, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 980, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_2180_0", + "timepoint": 2180, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2180, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; assistant,operating table,closeto; needle,patient,touching; head surgeon,needle,holding; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; syringe,needle,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_62_0", + "timepoint": 62, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 62, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,curette,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_124_0", + "timepoint": 124, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 124, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,scalpel,holding; assistant,forceps,holding; head surgeon,patient,cutting; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_802_0", + "timepoint": 802, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 802, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,body marker,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_19_0", + "timepoint": 19, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 19, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2707_0", + "timepoint": 2707, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2707, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; syringe,needle,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_792_0", + "timepoint": 792, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 792, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,microscope,manipulating; circulator,microscope,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1422_0", + "timepoint": 1422, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1422, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,instrument table,closeto; assistant,body marker,holding; assistant,instrument table,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_79_0", + "timepoint": 79, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 79, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2593_0", + "timepoint": 2593, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2593, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; syringe,needle,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,needle,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_870_0", + "timepoint": 870, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 870, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1044_0", + "timepoint": 1044, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1044, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_321_0", + "timepoint": 321, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 321, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_935_0", + "timepoint": 935, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 935, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_2326_0", + "timepoint": 2326, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2326, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_1_1_1169_0", + "timepoint": 1169, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1169, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating room,entering; assistant,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_2_2_494_0", + "timepoint": 494, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 494, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1904_0", + "timepoint": 1904, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1904, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound screen,looking; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2436_0", + "timepoint": 2436, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2436, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; needle,patient,touching; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; syringe,needle,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_638_0", + "timepoint": 638, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 638, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2897_0", + "timepoint": 2897, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2897, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,operating table,closeto; assistant,syringe,removing; head surgeon,vertebrae,scanning; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_242_0", + "timepoint": 242, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 242, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_286_0", + "timepoint": 286, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 286, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,operating table,manipulating; patient,operating table,lyingon; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_3092_0", + "timepoint": 3092, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3092, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_828_0", + "timepoint": 828, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 828, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,body marker,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2263_0", + "timepoint": 2263, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2263, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,syringe,removing; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; syringe,needle,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_747_0", + "timepoint": 747, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 747, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1812_0", + "timepoint": 1812, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1812, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,inserting; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1126_0", + "timepoint": 1126, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1126, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2457_0", + "timepoint": 2457, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2457, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; needle,patient,touching; head surgeon,needle,holding; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2415_0", + "timepoint": 2415, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2415, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; needle,patient,touching; patient,operating table,lyingon; head surgeon,needle,holding; assistant,syringe,holding; assistant,instrument table,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_2068_0", + "timepoint": 2068, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2068, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; forceps,cotton,touching; assistant,forceps,holding; assistant,patient,disinfection; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2341_0", + "timepoint": 2341, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2341, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2101_0", + "timepoint": 2101, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2101, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; head surgeon,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2889_0", + "timepoint": 2889, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2889, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,syringe,removing; syringe,needle,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_2356_0", + "timepoint": 2356, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2356, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,removing; patient,operating table,lyingon; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3226_0", + "timepoint": 3226, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3226, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; assistant,tissue paper,applying; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_631_0", + "timepoint": 631, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 631, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,curette,holding; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_3206_0", + "timepoint": 3206, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3206, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2479_0", + "timepoint": 2479, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2479, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound machine,looking; assistant,instrument table,closeto; assistant,instrument table,looking; assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_296_0", + "timepoint": 296, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 296, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; anesthetist,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_1928_0", + "timepoint": 1928, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1928, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,instrument table,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_803_0", + "timepoint": 803, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 803, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1353_0", + "timepoint": 1353, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1353, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_624_0", + "timepoint": 624, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 624, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,microscope screen,looking; anesthetist,operating table,closeto; assistant,microscope screen,looking; head surgeon,microscope controller,controlling; patient,operating table,lyingon; anesthetist,health monitor,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2689_0", + "timepoint": 2689, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2689, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1164_0", + "timepoint": 1164, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1164, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,body marker,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1579_0", + "timepoint": 1579, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1579, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2361_0", + "timepoint": 2361, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2361, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,needle,removing; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_1174_0", + "timepoint": 1174, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1174, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_455_0", + "timepoint": 455, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 455, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_209_0", + "timepoint": 209, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 209, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2413_0", + "timepoint": 2413, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2413, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound machine,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; needle,patient,touching; assistant,needle,removing " + } + ] + }, + { + "id": "Ultrasound_3_4_2956_0", + "timepoint": 2956, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2956, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,tissue paper,applying; tissue paper,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1293_0", + "timepoint": 1293, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1293, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1161_0", + "timepoint": 1161, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1161, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1560_0", + "timepoint": 1560, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1560, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2484_0", + "timepoint": 2484, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2484, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; syringe,needle,touching; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,needle,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1606_0", + "timepoint": 1606, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1606, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1162_0", + "timepoint": 1162, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1162, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating room,entering; assistant,operating room,entering; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_509_0", + "timepoint": 509, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 509, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_458_0", + "timepoint": 458, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 458, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,manipulating; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1041_0", + "timepoint": 1041, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1041, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_586_0", + "timepoint": 586, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 586, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_674_0", + "timepoint": 674, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 674, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,controlling; assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,microscope screen,looking; head surgeon,operating table,closeto; head surgeon,microscope screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1353_0", + "timepoint": 1353, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1353, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,body marker,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2280_0", + "timepoint": 2280, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2280, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_573_0", + "timepoint": 573, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 573, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,microscope controller,controlling; assistant,operating table,closeto; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_917_0", + "timepoint": 917, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 917, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; patient,operating table,lyingon; cotton,patient,touching; assistant,operating table,closeto; assistant,antiseptic,holding; assistant,patient,disinfection; forceps,cotton,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_407_0", + "timepoint": 407, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 407, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; circulator,patient,closeto; patient,operating table,touching " + } + ] + }, + { + "id": "MISS_3_6_1169_0", + "timepoint": 1169, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1169, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "MISS_3_6_1180_0", + "timepoint": 1180, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1180, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_557_0", + "timepoint": 557, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 557, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1263_0", + "timepoint": 1263, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1263, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; body marker,patient,touching; assistant,body marker,applying; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_89_0", + "timepoint": 89, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 89, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1237_0", + "timepoint": 1237, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1237, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1153_0", + "timepoint": 1153, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1153, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_849_0", + "timepoint": 849, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 849, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,body marker,applying; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1899_0", + "timepoint": 1899, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1899, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; syringe,needle,touching; assistant,syringe,injecting; head surgeon,needle,holding; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2219_0", + "timepoint": 2219, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2219, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; patient,operating table,lyingon; assistant,needle,holding; syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_620_0", + "timepoint": 620, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 620, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_1291_0", + "timepoint": 1291, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1291, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,antiseptic,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_472_0", + "timepoint": 472, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 472, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,forceps,holding; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; head surgeon,scalpel,holding; assistant,operating table,closeto; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_203_0", + "timepoint": 203, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 203, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,looking; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2227_0", + "timepoint": 2227, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2227, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,holding; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_93_0", + "timepoint": 93, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 93, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,ultrasound gel,holding; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_123_0", + "timepoint": 123, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 123, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2715_0", + "timepoint": 2715, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2715, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,needle,holding; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2756_0", + "timepoint": 2756, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2756, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; needle,patient,touching; head surgeon,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2725_0", + "timepoint": 2725, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2725, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,instrument table,looking; needle,patient,touching; head surgeon,needle,holding; assistant,syringe,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_14_0", + "timepoint": 14, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 14, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_100_0", + "timepoint": 100, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 100, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,patient,manipulating; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2054_0", + "timepoint": 2054, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2054, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; assistant,needle,holding; assistant,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,syringe,injecting; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2338_0", + "timepoint": 2338, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2338, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; syringe,needle,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_477_0", + "timepoint": 477, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 477, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,scalpel,holding; curette,patient,touching; scalpel,patient,touching; head surgeon,curette,holding; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,forceps,holding; head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1056_0", + "timepoint": 1056, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1056, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_385_0", + "timepoint": 385, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 385, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,aspirating; forceps,cotton,touching; anesthetist,operating table,closeto; assistant,operating table,closeto; cotton,patient,touching; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1094_0", + "timepoint": 1094, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1094, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_337_0", + "timepoint": 337, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 337, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_3003_0", + "timepoint": 3003, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3003, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; tissue paper,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2281_0", + "timepoint": 2281, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2281, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; needle,patient,touching; assistant,syringe,holding " + } + ] + }, + { + "id": "MISS_3_6_768_0", + "timepoint": 768, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 768, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; assistant,instrument table,closeto; assistant,instruments,disinfection; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,anesthetist,applying; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2365_0", + "timepoint": 2365, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2365, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1933_0", + "timepoint": 1933, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1933, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,needle,holding; assistant,instrument table,looking; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1011_0", + "timepoint": 1011, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1011, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3268_0", + "timepoint": 3268, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3268, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; tissue paper,patient,touching; assistant,tissue paper,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_124_0", + "timepoint": 124, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 124, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_153_0", + "timepoint": 153, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 153, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,manipulating; head surgeon,ultrasound machine,looking; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1958_0", + "timepoint": 1958, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1958, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_3_6_129_0", + "timepoint": 129, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 129, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; curette,patient,touching; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2850_0", + "timepoint": 2850, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2850, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2941_0", + "timepoint": 2941, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2941, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,syringe,holding; assistant,needle,removing; head surgeon,patient,scanning; assistant,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1459_0", + "timepoint": 1459, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1459, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2458_0", + "timepoint": 2458, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2458, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,instrument table,closeto; needle,patient,touching; patient,operating table,lyingon; head surgeon,needle,holding; assistant,syringe,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2636_0", + "timepoint": 2636, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2636, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,needle,holding; assistant,syringe,injecting; needle,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_256_0", + "timepoint": 256, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 256, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; patient,operating table,lyingon; forceps,cotton,touching; scalpel,patient,touching; assistant,forceps,holding; head surgeon,operating table,closeto; head surgeon,patient,cutting; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_761_0", + "timepoint": 761, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 761, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_882_0", + "timepoint": 882, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 882, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; body marker,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,body marker,applying " + } + ] + }, + { + "id": "Ultrasound_3_4_2949_0", + "timepoint": 2949, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2949, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying; tissue paper,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1778_0", + "timepoint": 1778, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1778, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,instrument table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1538_0", + "timepoint": 1538, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1538, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2040_0", + "timepoint": 2040, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2040, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; syringe,needle,touching; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1533_0", + "timepoint": 1533, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1533, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,inserting; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,ultrasound probe,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_188_0", + "timepoint": 188, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 188, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2404_0", + "timepoint": 2404, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2404, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; needle,patient,touching; assistant,syringe,injecting; syringe,needle,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_1480_0", + "timepoint": 1480, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1480, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope controller,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1953_0", + "timepoint": 1953, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1953, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1175_0", + "timepoint": 1175, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1175, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2154_0", + "timepoint": 2154, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2154, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_560_0", + "timepoint": 560, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 560, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_585_0", + "timepoint": 585, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 585, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_677_0", + "timepoint": 677, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 677, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,microscope screen,looking; assistant,microscope screen,looking; head surgeon,microscope controller,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_313_0", + "timepoint": 313, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 313, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; patient,operating room,entering " + } + ] + }, + { + "id": "MISS_1_1_1435_0", + "timepoint": 1435, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1435, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,patient,anaesthetising " + } + ] + }, + { + "id": "MISS_2_2_771_0", + "timepoint": 771, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 771, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,health monitor,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_1462_0", + "timepoint": 1462, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1462, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,patient,anaesthetising; head surgeon,microscope controller,manipulating; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3036_0", + "timepoint": 3036, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3036, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; head surgeon,tissue paper,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1793_0", + "timepoint": 1793, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1793, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; assistant,needle,inserting; assistant,tissue mark,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1265_0", + "timepoint": 1265, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1265, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding; anesthetist,patient,manipulating; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1874_0", + "timepoint": 1874, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1874, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; ultrasound probe,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_597_0", + "timepoint": 597, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 597, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; curette,patient,touching; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,health monitor,looking; scalpel,patient,touching; head surgeon,curette,holding; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "MISS_3_6_956_0", + "timepoint": 956, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 956, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,aspirating; anesthetist,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; forceps,cotton,touching; cotton,patient,touching; head surgeon,operating table,closeto; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_290_0", + "timepoint": 290, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 290, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound gel,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2435_0", + "timepoint": 2435, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2435, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,patient,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_143_0", + "timepoint": 143, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 143, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1789_0", + "timepoint": 1789, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1789, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,operating table,closeto; assistant,needle,inserting; assistant,tissue mark,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1048_0", + "timepoint": 1048, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1048, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1025_0", + "timepoint": 1025, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1025, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; forceps,cotton,touching; assistant,operating table,closeto; assistant,patient,aspirating; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,forceps,holding; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; cotton,patient,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1704_0", + "timepoint": 1704, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1704, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_3158_0", + "timepoint": 3158, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3158, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,tissue paper,applying; head surgeon,operating table,closeto; patient,operating table,lyingon; tissue paper,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_947_0", + "timepoint": 947, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 947, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1392_0", + "timepoint": 1392, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1392, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1384_0", + "timepoint": 1384, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1384, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; assistant,needle,removing; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2960_0", + "timepoint": 2960, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2960, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying; tissue paper,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1219_0", + "timepoint": 1219, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1219, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; circulator,ultrasound machine,manipulating; head surgeon,operating table,closeto; forceps,cotton,touching; circulator,ultrasound machine,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_1_1_1661_0", + "timepoint": 1661, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1661, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_616_0", + "timepoint": 616, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 616, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1368_0", + "timepoint": 1368, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1368, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,body marker,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_1048_0", + "timepoint": 1048, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1048, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding; forceps,cotton,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2622_0", + "timepoint": 2622, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2622, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,operating table,closeto; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_1_1_314_0", + "timepoint": 314, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 314, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_532_0", + "timepoint": 532, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 532, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating room,entering; circulator,assistant,closeto " + } + ] + }, + { + "id": "MISS_3_6_963_0", + "timepoint": 963, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 963, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; forceps,cotton,touching; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,patient,aspirating; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_755_0", + "timepoint": 755, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 755, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,instruments,disinfection; assistant,instrument table,closeto; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; assistant,anesthetist,applying; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1900_0", + "timepoint": 1900, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1900, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_2_2_731_0", + "timepoint": 731, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 731, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,microscope screen,looking; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_259_0", + "timepoint": 259, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 259, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_81_0", + "timepoint": 81, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 81, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2203_0", + "timepoint": 2203, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2203, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; syringe,needle,touching; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; head surgeon,needle,holding; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1176_0", + "timepoint": 1176, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1176, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2018_0", + "timepoint": 2018, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2018, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; assistant,patient,disinfection; assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_959_0", + "timepoint": 959, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 959, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,patient,disinfection; assistant,antiseptic,holding; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_715_0", + "timepoint": 715, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 715, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_688_0", + "timepoint": 688, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 688, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,instrument table,looking; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_765_0", + "timepoint": 765, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 765, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1226_0", + "timepoint": 1226, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1226, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,patient,manipulating; anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2604_0", + "timepoint": 2604, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2604, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,tissue paper,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_425_0", + "timepoint": 425, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 425, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,instrument table,closeto; assistant,ultrasound gel,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_3_6_1044_0", + "timepoint": 1044, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1044, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,microscope eye,looking; cotton,patient,touching; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,health monitor,looking; anesthetist,operating table,closeto; assistant,patient,aspirating; forceps,cotton,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_385_0", + "timepoint": 385, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 385, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_973_0", + "timepoint": 973, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 973, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1609_0", + "timepoint": 1609, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1609, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; head surgeon,operating table,closeto; needle,patient,touching; assistant,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_131_0", + "timepoint": 131, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 131, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,curette,holding; curette,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2131_0", + "timepoint": 2131, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2131, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; syringe,needle,touching; assistant,needle,holding; assistant,syringe,injecting; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2107_0", + "timepoint": 2107, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2107, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; assistant,operating table,closeto; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_422_0", + "timepoint": 422, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 422, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,instrument table,closeto; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2218_0", + "timepoint": 2218, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2218, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1061_0", + "timepoint": 1061, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1061, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,body marker,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_788_0", + "timepoint": 788, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 788, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,body marker,holding; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2390_0", + "timepoint": 2390, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2390, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3184_0", + "timepoint": 3184, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3184, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; patient,operating table,lyingon; assistant,tissue paper,applying; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3082_0", + "timepoint": 3082, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3082, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2298_0", + "timepoint": 2298, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2298, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2694_0", + "timepoint": 2694, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2694, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,tissue paper,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2284_0", + "timepoint": 2284, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2284, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2667_0", + "timepoint": 2667, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2667, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; syringe,needle,touching; assistant,needle,holding; ultrasound probe,patient,touching; assistant,syringe,injecting; needle,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_2_2_241_0", + "timepoint": 241, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 241, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,scalpel,holding; assistant,operating table,closeto; assistant,forceps,holding; anesthetist,operating table,closeto; anesthetist,health monitor,looking; head surgeon,patient,cutting; scalpel,patient,touching; anesthetist,anesthesia equipment,controlling; forceps,cotton,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2385_0", + "timepoint": 2385, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2385, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_9_0", + "timepoint": 9, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 9, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1713_0", + "timepoint": 1713, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1713, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2064_0", + "timepoint": 2064, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2064, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_2_2_630_0", + "timepoint": 630, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 630, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,microscope screen,looking; assistant,microscope screen,looking; head surgeon,microscope controller,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1836_0", + "timepoint": 1836, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1836, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,needle,removing; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2727_0", + "timepoint": 2727, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2727, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; syringe,needle,touching; assistant,syringe,injecting; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_473_0", + "timepoint": 473, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 473, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,instrument table,closeto; anesthetist,health monitor,looking; assistant,scalpel,holding " + } + ] + }, + { + "id": "MISS_1_1_2187_0", + "timepoint": 2187, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2187, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; forceps,cotton,touching; head surgeon,operating table,closeto; cotton,patient,touching; assistant,forceps,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_5_6_85_0", + "timepoint": 85, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 85, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,ultrasound gel,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_918_0", + "timepoint": 918, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 918, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "MISS_1_1_1493_0", + "timepoint": 1493, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1493, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope controller,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_524_0", + "timepoint": 524, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 524, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; curette,patient,touching; scalpel,patient,touching; head surgeon,operating table,closeto; head surgeon,curette,holding; head surgeon,scalpel,holding; head surgeon,herbal disk,removing; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1786_0", + "timepoint": 1786, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1786, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2184_0", + "timepoint": 2184, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2184, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; assistant,operating table,closeto; needle,patient,touching; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1487_0", + "timepoint": 1487, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1487, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_2229_0", + "timepoint": 2229, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2229, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,operating table,closeto; cotton,patient,touching; anesthetist,operating table,closeto; assistant,patient,disinfection; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_1_1_427_0", + "timepoint": 427, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 427, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,operating table,manipulating; anesthetist,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_707_0", + "timepoint": 707, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 707, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_349_0", + "timepoint": 349, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 349, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,applying " + } + ] + }, + { + "id": "Ultrasound_2_2_1165_0", + "timepoint": 1165, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1165, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,antiseptic,holding; circulator,ultrasound machine,closeto; patient,operating table,lyingon; circulator,ultrasound machine,manipulating; forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_555_0", + "timepoint": 555, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 555, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_274_0", + "timepoint": 274, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 274, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; patient,operating table,lyingon; head surgeon,herbal disk,removing; head surgeon,microscope eye,looking; head surgeon,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,curette,holding; forceps,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2761_0", + "timepoint": 2761, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2761, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; needle,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2647_0", + "timepoint": 2647, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2647, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1090_0", + "timepoint": 1090, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1090, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,health monitor,looking; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; assistant,forceps,holding; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_597_0", + "timepoint": 597, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 597, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1039_0", + "timepoint": 1039, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1039, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_1040_0", + "timepoint": 1040, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1040, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,health monitor,looking; assistant,forceps,holding; forceps,cotton,touching; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,patient,aspirating " + } + ] + }, + { + "id": "Ultrasound_3_4_446_0", + "timepoint": 446, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 446, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_59_0", + "timepoint": 59, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 59, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1373_0", + "timepoint": 1373, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1373, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_190_0", + "timepoint": 190, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 190, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_448_0", + "timepoint": 448, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 448, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,scalpel,holding; patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,curette,holding; curette,patient,touching; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1044_0", + "timepoint": 1044, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1044, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1627_0", + "timepoint": 1627, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1627, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,tissue mark,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1893_0", + "timepoint": 1893, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1893, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,operating table,closeto; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_1_1_2305_0", + "timepoint": 2305, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2305, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,forceps,holding; assistant,instrument table,closeto; patient,operating table,lyingon; forceps,cotton,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1436_0", + "timepoint": 1436, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1436, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2063_0", + "timepoint": 2063, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2063, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,needle,holding; assistant,needle,holding; assistant,syringe,injecting; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1135_0", + "timepoint": 1135, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1135, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_624_0", + "timepoint": 624, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 624, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2070_0", + "timepoint": 2070, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2070, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_2477_0", + "timepoint": 2477, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2477, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound machine,looking; patient,operating table,lyingon; assistant,instrument table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_403_0", + "timepoint": 403, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 403, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_339_0", + "timepoint": 339, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 339, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; patient,operating table,lyingon; anesthetist,operating table,closeto; curette,patient,touching; assistant,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2143_0", + "timepoint": 2143, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2143, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; patient,operating table,lyingon; syringe,needle,touching; head surgeon,operating table,closeto; needle,patient,touching; assistant,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1957_0", + "timepoint": 1957, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1957, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,antiseptic,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1287_0", + "timepoint": 1287, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1287, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1185_0", + "timepoint": 1185, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1185, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,body marker,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_646_0", + "timepoint": 646, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 646, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,controlling; head surgeon,microscope screen,looking; patient,operating table,lyingon; assistant,microscope screen,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2403_0", + "timepoint": 2403, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2403, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,instrument table,looking; head surgeon,needle,holding; assistant,syringe,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_13_0", + "timepoint": 13, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 13, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_161_0", + "timepoint": 161, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 161, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound machine,looking; assistant,patient,manipulating; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1959_0", + "timepoint": 1959, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1959, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,syringe,holding; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1649_0", + "timepoint": 1649, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1649, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,needle,inserting; assistant,operating table,closeto; assistant,tissue mark,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1400_0", + "timepoint": 1400, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1400, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_680_0", + "timepoint": 680, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 680, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1129_0", + "timepoint": 1129, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1129, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; patient,operating table,lyingon; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2073_0", + "timepoint": 2073, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2073, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1833_0", + "timepoint": 1833, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1833, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; syringe,needle,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2633_0", + "timepoint": 2633, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2633, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,tissue paper,applying; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_917_0", + "timepoint": 917, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 917, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; forceps,cotton,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2422_0", + "timepoint": 2422, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2422, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; assistant,instrument table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,instrument table,looking; head surgeon,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1281_0", + "timepoint": 1281, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1281, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto; anesthetist,patient,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_6_0", + "timepoint": 6, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 6, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1430_0", + "timepoint": 1430, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1430, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,patient,anaesthetising; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_2003_0", + "timepoint": 2003, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2003, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,syringe,holding; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2918_0", + "timepoint": 2918, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2918, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,syringe,holding; assistant,needle,holding; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_461_0", + "timepoint": 461, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 461, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_80_0", + "timepoint": 80, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 80, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_984_0", + "timepoint": 984, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 984, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,body marker,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_945_0", + "timepoint": 945, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 945, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_964_0", + "timepoint": 964, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 964, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,body marker,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2333_0", + "timepoint": 2333, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2333, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,removing; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1547_0", + "timepoint": 1547, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1547, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1298_0", + "timepoint": 1298, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1298, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,needle,inserting; needle,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_469_0", + "timepoint": 469, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 469, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; anesthetist,health monitor,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,scalpel,holding " + } + ] + }, + { + "id": "MISS_1_1_2157_0", + "timepoint": 2157, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2157, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; cotton,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2172_0", + "timepoint": 2172, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2172, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; needle,patient,touching; syringe,needle,touching; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_3_6_383_0", + "timepoint": 383, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 383, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; curette,patient,touching; assistant,forceps,holding; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,forceps,holding; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1970_0", + "timepoint": 1970, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1970, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; syringe,needle,touching; assistant,operating table,closeto; needle,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1988_0", + "timepoint": 1988, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1988, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,syringe,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2738_0", + "timepoint": 2738, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2738, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,instrument table,looking; needle,patient,touching; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_741_0", + "timepoint": 741, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 741, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,microscope screen,looking; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_471_0", + "timepoint": 471, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 471, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,scalpel,holding; patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_612_0", + "timepoint": 612, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 612, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_360_0", + "timepoint": 360, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 360, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; circulator,patient,looking; circulator,patient,closeto; patient,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2292_0", + "timepoint": 2292, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2292, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,needle,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1215_0", + "timepoint": 1215, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1215, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_661_0", + "timepoint": 661, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 661, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope controller,controlling; head surgeon,microscope screen,looking; assistant,microscope screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_54_0", + "timepoint": 54, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 54, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound machine,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1057_0", + "timepoint": 1057, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1057, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_654_0", + "timepoint": 654, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 654, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1260_0", + "timepoint": 1260, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1260, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_654_0", + "timepoint": 654, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 654, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,microscope screen,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope screen,looking; head surgeon,microscope controller,controlling; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_327_0", + "timepoint": 327, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 327, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_249_0", + "timepoint": 249, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 249, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_311_0", + "timepoint": 311, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 311, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,patient,closeto " + } + ] + }, + { + "id": "MISS_3_6_54_0", + "timepoint": 54, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 54, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,curette,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1212_0", + "timepoint": 1212, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1212, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_757_0", + "timepoint": 757, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 757, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1801_0", + "timepoint": 1801, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1801, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,antiseptic,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_622_0", + "timepoint": 622, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 622, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_408_0", + "timepoint": 408, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 408, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,looking; head surgeon,vertebrae,scanning; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2565_0", + "timepoint": 2565, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2565, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "MISS_1_1_368_0", + "timepoint": 368, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 368, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_3219_0", + "timepoint": 3219, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3219, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; tissue paper,patient,touching; assistant,tissue paper,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1195_0", + "timepoint": 1195, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1195, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; circulator,ultrasound machine,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; patient,operating table,lyingon; circulator,ultrasound machine,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_981_0", + "timepoint": 981, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 981, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,patient,aspirating; cotton,patient,touching; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_695_0", + "timepoint": 695, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 695, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2380_0", + "timepoint": 2380, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2380, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; needle,patient,touching; assistant,needle,removing; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1493_0", + "timepoint": 1493, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1493, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_237_0", + "timepoint": 237, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 237, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; head surgeon,forceps,holding; curette,patient,touching; anesthetist,operating table,closeto; head surgeon,herbal disk,removing; assistant,operating table,closeto; head surgeon,curette,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_679_0", + "timepoint": 679, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 679, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "MISS_3_6_212_0", + "timepoint": 212, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 212, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,forceps,holding; curette,patient,touching; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,curette,holding; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1253_0", + "timepoint": 1253, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1253, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_666_0", + "timepoint": 666, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 666, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,looking; head surgeon,microscope eye,looking; assistant,instrument table,closeto; assistant,curette,holding; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2929_0", + "timepoint": 2929, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2929, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_473_0", + "timepoint": 473, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 473, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1694_0", + "timepoint": 1694, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1694, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,syringe,holding; assistant,instrument table,closeto; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_763_0", + "timepoint": 763, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 763, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,microscope screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1175_0", + "timepoint": 1175, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1175, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,inserting; assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1279_0", + "timepoint": 1279, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1279, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,forceps,holding; assistant,antiseptic,holding; head surgeon,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1257_0", + "timepoint": 1257, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1257, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; anesthetist,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_800_0", + "timepoint": 800, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 800, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_585_0", + "timepoint": 585, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 585, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_951_0", + "timepoint": 951, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 951, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; anesthetist,operating table,closeto; circulator,microscope,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1343_0", + "timepoint": 1343, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1343, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_465_0", + "timepoint": 465, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 465, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_300_0", + "timepoint": 300, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 300, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1900_0", + "timepoint": 1900, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1900, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,antiseptic,applying; assistant,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_2_2_18_0", + "timepoint": 18, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 18, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_903_0", + "timepoint": 903, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 903, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,body marker,applying; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; body marker,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_351_0", + "timepoint": 351, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 351, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; curette,patient,touching; head surgeon,curette,holding; head surgeon,herbal disk,removing; anesthetist,operating table,closeto; forceps,patient,touching; head surgeon,forceps,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_643_0", + "timepoint": 643, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 643, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_974_0", + "timepoint": 974, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 974, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching; assistant,antiseptic,holding; cotton,patient,touching; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_2_2_490_0", + "timepoint": 490, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 490, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1434_0", + "timepoint": 1434, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1434, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_149_0", + "timepoint": 149, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 149, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1070_0", + "timepoint": 1070, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1070, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2630_0", + "timepoint": 2630, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2630, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; assistant,operating table,closeto; assistant,needle,holding; needle,patient,touching; syringe,needle,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2079_0", + "timepoint": 2079, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2079, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,needle,holding; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2624_0", + "timepoint": 2624, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2624, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,tissue paper,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_52_0", + "timepoint": 52, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 52, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,curette,holding; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1026_0", + "timepoint": 1026, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1026, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; forceps,cotton,touching; assistant,patient,aspirating; anesthetist,operating table,closeto; anesthetist,health monitor,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1593_0", + "timepoint": 1593, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1593, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_822_0", + "timepoint": 822, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 822, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1375_0", + "timepoint": 1375, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1375, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,body marker,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_144_0", + "timepoint": 144, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 144, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,scissors,looking; circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_262_0", + "timepoint": 262, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 262, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_440_0", + "timepoint": 440, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 440, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating room,entering; circulator,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1364_0", + "timepoint": 1364, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1364, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,removing; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2725_0", + "timepoint": 2725, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2725, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,bin,closeto " + } + ] + }, + { + "id": "MISS_3_6_581_0", + "timepoint": 581, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 581, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; patient,operating table,lyingon; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; scalpel,patient,touching; anesthetist,operating table,closeto; curette,patient,touching; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "MISS_1_1_1737_0", + "timepoint": 1737, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1737, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2745_0", + "timepoint": 2745, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2745, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; syringe,needle,touching; needle,patient,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_595_0", + "timepoint": 595, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 595, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_228_0", + "timepoint": 228, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 228, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,instrument table,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1938_0", + "timepoint": 1938, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1938, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; syringe,needle,touching; patient,operating table,lyingon; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,needle,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2625_0", + "timepoint": 2625, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2625, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1318_0", + "timepoint": 1318, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1318, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,patient,manipulating; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; assistant,instrument table,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_768_0", + "timepoint": 768, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 768, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_1053_0", + "timepoint": 1053, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1053, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2086_0", + "timepoint": 2086, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2086, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_2_2_639_0", + "timepoint": 639, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 639, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,microscope screen,looking; assistant,operating table,closeto; head surgeon,microscope controller,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2017_0", + "timepoint": 2017, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2017, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; assistant,syringe,holding; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1843_0", + "timepoint": 1843, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1843, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,patient,disinfection; assistant,operating table,closeto; assistant,antiseptic,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_235_0", + "timepoint": 235, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 235, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1973_0", + "timepoint": 1973, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1973, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,needle,inserting; assistant,operating table,closeto; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2081_0", + "timepoint": 2081, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2081, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; head surgeon,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_5_6_2567_0", + "timepoint": 2567, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2567, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; needle,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_137_0", + "timepoint": 137, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 137, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,cutting; anesthetist,operating table,closeto; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,scalpel,holding; scalpel,patient,touching; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_1_1_527_0", + "timepoint": 527, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 527, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,operating table,manipulating; anesthetist,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1326_0", + "timepoint": 1326, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1326, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_984_0", + "timepoint": 984, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 984, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,instrument table,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1893_0", + "timepoint": 1893, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1893, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_212_0", + "timepoint": 212, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 212, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling; circulator,syringe,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_944_0", + "timepoint": 944, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 944, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_600_0", + "timepoint": 600, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 600, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1514_0", + "timepoint": 1514, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1514, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2827_0", + "timepoint": 2827, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2827, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; syringe,needle,touching; assistant,syringe,removing; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2851_0", + "timepoint": 2851, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2851, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; patient,operating table,lyingon; assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_708_0", + "timepoint": 708, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 708, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,microscope screen,looking; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2130_0", + "timepoint": 2130, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2130, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2772_0", + "timepoint": 2772, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2772, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_773_0", + "timepoint": 773, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 773, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,health monitor,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_2_2_585_0", + "timepoint": 585, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 585, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,controlling " + } + ] + }, + { + "id": "MISS_3_6_883_0", + "timepoint": 883, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 883, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_1686_0", + "timepoint": 1686, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1686, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2425_0", + "timepoint": 2425, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2425, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; needle,patient,touching; assistant,instrument table,looking; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1977_0", + "timepoint": 1977, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1977, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,syringe,holding " + } + ] + }, + { + "id": "MISS_1_1_1563_0", + "timepoint": 1563, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1563, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,patient,anaesthetising; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1038_0", + "timepoint": 1038, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1038, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_1809_0", + "timepoint": 1809, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1809, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_625_0", + "timepoint": 625, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 625, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1330_0", + "timepoint": 1330, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1330, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,inserting; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1190_0", + "timepoint": 1190, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1190, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,body marker,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_897_0", + "timepoint": 897, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 897, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,body marker,applying; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; body marker,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3256_0", + "timepoint": 3256, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3256, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; assistant,operating table,closeto; assistant,tissue paper,applying; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_293_0", + "timepoint": 293, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 293, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,scalpel,holding; patient,operating table,lyingon; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_21_0", + "timepoint": 21, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 21, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_3041_0", + "timepoint": 3041, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3041, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1133_0", + "timepoint": 1133, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1133, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; assistant,instrument table,closeto; anesthetist,health monitor,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_683_0", + "timepoint": 683, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 683, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_923_0", + "timepoint": 923, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 923, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,microscope,manipulating; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1910_0", + "timepoint": 1910, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1910, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; syringe,needle,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2236_0", + "timepoint": 2236, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2236, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1116_0", + "timepoint": 1116, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1116, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_695_0", + "timepoint": 695, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 695, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_728_0", + "timepoint": 728, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 728, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; assistant,antiseptic,holding; assistant,instrument table,looking; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_75_0", + "timepoint": 75, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 75, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2470_0", + "timepoint": 2470, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2470, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; head surgeon,vertebrae,scanning; head surgeon,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,instrument table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_798_0", + "timepoint": 798, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 798, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_296_0", + "timepoint": 296, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 296, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_298_0", + "timepoint": 298, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 298, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_1091_0", + "timepoint": 1091, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1091, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2184_0", + "timepoint": 2184, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2184, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; needle,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; syringe,needle,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_270_0", + "timepoint": 270, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 270, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_2660_0", + "timepoint": 2660, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2660, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,needle,holding; assistant,syringe,injecting; head surgeon,needle,holding; syringe,needle,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_774_0", + "timepoint": 774, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 774, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_2165_0", + "timepoint": 2165, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2165, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; forceps,cotton,touching; assistant,patient,disinfection; assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_13_0", + "timepoint": 13, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 13, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_615_0", + "timepoint": 615, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 615, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_200_0", + "timepoint": 200, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 200, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_154_0", + "timepoint": 154, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 154, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2776_0", + "timepoint": 2776, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2776, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,operating table,closeto; assistant,needle,removing; head surgeon,ultrasound probe,holding; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_3035_0", + "timepoint": 3035, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3035, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; tissue paper,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1313_0", + "timepoint": 1313, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1313, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,inserting; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_954_0", + "timepoint": 954, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 954, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1659_0", + "timepoint": 1659, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1659, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound screen,looking; needle,patient,touching; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_231_0", + "timepoint": 231, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 231, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2566_0", + "timepoint": 2566, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2566, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,needle,holding; ultrasound probe,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_926_0", + "timepoint": 926, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 926, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_193_0", + "timepoint": 193, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 193, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_933_0", + "timepoint": 933, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 933, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; forceps,cotton,touching; assistant,operating table,closeto; assistant,antiseptic,holding; head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2742_0", + "timepoint": 2742, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2742, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3059_0", + "timepoint": 3059, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3059, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; assistant,tissue paper,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1160_0", + "timepoint": 1160, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1160, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; patient,operating table,lyingon; assistant,antiseptic,holding; assistant,forceps,holding; head surgeon,operating table,closeto; circulator,ultrasound machine,manipulating; circulator,ultrasound machine,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1022_0", + "timepoint": 1022, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1022, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_890_0", + "timepoint": 890, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 890, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_288_0", + "timepoint": 288, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 288, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,looking; patient,operating room,entering " + } + ] + }, + { + "id": "MISS_3_6_356_0", + "timepoint": 356, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 356, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; forceps,patient,touching; curette,patient,touching; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; head surgeon,forceps,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "Ultrasound_3_4_2714_0", + "timepoint": 2714, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2714, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding; needle,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_921_0", + "timepoint": 921, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 921, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,microscope,manipulating; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "MISS_1_1_1424_0", + "timepoint": 1424, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1424, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,microscope controller,manipulating; anesthetist,patient,anaesthetising; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1110_0", + "timepoint": 1110, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1110, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; assistant,ultrasound probe,looking; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1437_0", + "timepoint": 1437, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1437, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1201_0", + "timepoint": 1201, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1201, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating room,entering; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_764_0", + "timepoint": 764, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 764, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_46_0", + "timepoint": 46, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 46, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound machine,looking; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_244_0", + "timepoint": 244, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 244, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,patient,touching; head surgeon,forceps,holding; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,health monitor,looking; anesthetist,operating table,closeto; curette,patient,touching; head surgeon,herbal disk,removing; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2737_0", + "timepoint": 2737, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2737, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1885_0", + "timepoint": 1885, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1885, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3250_0", + "timepoint": 3250, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3250, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; tissue paper,patient,touching; assistant,tissue paper,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_431_0", + "timepoint": 431, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 431, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2742_0", + "timepoint": 2742, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2742, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; syringe,needle,touching; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2440_0", + "timepoint": 2440, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2440, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; head surgeon,ultrasound machine,looking; needle,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2404_0", + "timepoint": 2404, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2404, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,removing; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; needle,patient,touching; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "MISS_1_1_2162_0", + "timepoint": 2162, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2162, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,patient,disinfection; head surgeon,operating table,closeto; cotton,patient,touching; forceps,cotton,touching; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2120_0", + "timepoint": 2120, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2120, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; needle,patient,touching; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_692_0", + "timepoint": 692, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 692, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; circulator,ultrasound machine,looking; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "MISS_2_2_566_0", + "timepoint": 566, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 566, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,controlling; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,health monitor,looking; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_942_0", + "timepoint": 942, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 942, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_645_0", + "timepoint": 645, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 645, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_735_0", + "timepoint": 735, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 735, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_2_2_55_0", + "timepoint": 55, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 55, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; assistant,scalpel,holding; patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1065_0", + "timepoint": 1065, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1065, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,looking; assistant,operating table,closeto; assistant,antiseptic,holding; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_631_0", + "timepoint": 631, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 631, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_938_0", + "timepoint": 938, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 938, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,body marker,holding; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1323_0", + "timepoint": 1323, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1323, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,antiseptic,holding; assistant,forceps,holding; forceps,cotton,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_371_0", + "timepoint": 371, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 371, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "MISS_1_1_587_0", + "timepoint": 587, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 587, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_595_0", + "timepoint": 595, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 595, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,microscope controller,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1283_0", + "timepoint": 1283, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1283, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,body marker,applying; head surgeon,ultrasound probe,controlling; body marker,patient,touching; assistant,operating table,closeto; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "MISS_1_1_2362_0", + "timepoint": 2362, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2362, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_950_0", + "timepoint": 950, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 950, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,assistant,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_1050_0", + "timepoint": 1050, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1050, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; assistant,forceps,holding; assistant,patient,aspirating; cotton,patient,touching; head surgeon,operating table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "MISS_1_1_622_0", + "timepoint": 622, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 622, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_633_0", + "timepoint": 633, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 633, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,patient,manipulating; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1516_0", + "timepoint": 1516, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1516, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; assistant,needle,inserting; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1531_0", + "timepoint": 1531, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1531, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1989_0", + "timepoint": 1989, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1989, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,syringe,holding; assistant,needle,inserting; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1736_0", + "timepoint": 1736, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1736, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,syringe,holding; assistant,needle,holding; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_162_0", + "timepoint": 162, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 162, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,curette,holding; assistant,operating table,closeto; curette,patient,touching; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1831_0", + "timepoint": 1831, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1831, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,needle,holding; assistant,operating table,closeto; syringe,needle,touching; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_4_5_2609_0", + "timepoint": 2609, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2609, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_592_0", + "timepoint": 592, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 592, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering; circulator,operating table,manipulating " + } + ] + }, + { + "id": "MISS_3_6_56_0", + "timepoint": 56, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 56, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; assistant,curette,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_3_6_80_0", + "timepoint": 80, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 80, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; head surgeon,curette,holding; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_454_0", + "timepoint": 454, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 454, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; assistant,scalpel,holding; anesthetist,operating table,closeto; head surgeon,forceps,holding; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,herbal disk,removing; curette,patient,touching; head surgeon,microscope eye,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1559_0", + "timepoint": 1559, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1559, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_1535_0", + "timepoint": 1535, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1535, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,manipulating; assistant,operating table,closeto; anesthetist,patient,anaesthetising; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1685_0", + "timepoint": 1685, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1685, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,inserting; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1040_0", + "timepoint": 1040, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1040, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; assistant,antiseptic,holding; assistant,forceps,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,patient,disinfection; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2566_0", + "timepoint": 2566, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2566, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; syringe,needle,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; needle,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_4_5_1992_0", + "timepoint": 1992, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1992, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching; syringe,needle,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1557_0", + "timepoint": 1557, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1557, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound probe,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_67_0", + "timepoint": 67, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 67, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound machine,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1826_0", + "timepoint": 1826, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1826, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,applying; anesthetist,operating table,closeto; assistant,patient,disinfection; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_791_0", + "timepoint": 791, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 791, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,microscope screen,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1444_0", + "timepoint": 1444, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1444, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_684_0", + "timepoint": 684, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 684, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1776_0", + "timepoint": 1776, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1776, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,needle,holding; assistant,syringe,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1538_0", + "timepoint": 1538, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1538, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; assistant,needle,inserting; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2432_0", + "timepoint": 2432, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2432, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; assistant,syringe,injecting; assistant,needle,holding; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_2197_0", + "timepoint": 2197, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2197, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,operating table,closeto; assistant,patient,disinfection; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_238_0", + "timepoint": 238, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 238, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1868_0", + "timepoint": 1868, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1868, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_589_0", + "timepoint": 589, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 589, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1432_0", + "timepoint": 1432, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1432, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,removing; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_567_0", + "timepoint": 567, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 567, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_2_2_705_0", + "timepoint": 705, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 705, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,microscope screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_884_0", + "timepoint": 884, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 884, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1232_0", + "timepoint": 1232, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1232, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,antiseptic,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_293_0", + "timepoint": 293, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 293, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering; circulator,operating table,manipulating; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_1378_0", + "timepoint": 1378, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1378, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,body marker,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1406_0", + "timepoint": 1406, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1406, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; needle,patient,touching; assistant,needle,removing " + } + ] + }, + { + "id": "MISS_1_1_1886_0", + "timepoint": 1886, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1886, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; assistant,operating table,closeto; assistant,antiseptic,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_284_0", + "timepoint": 284, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 284, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_67_0", + "timepoint": 67, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 67, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2325_0", + "timepoint": 2325, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2325, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_973_0", + "timepoint": 973, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 973, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,instrument table,closeto; assistant,instrument table,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2256_0", + "timepoint": 2256, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2256, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,removing; assistant,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1670_0", + "timepoint": 1670, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1670, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_64_0", + "timepoint": 64, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 64, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,scalpel,holding; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2255_0", + "timepoint": 2255, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2255, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,forceps,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,patient,disinfection; assistant,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_726_0", + "timepoint": 726, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 726, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; assistant,instrument table,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2863_0", + "timepoint": 2863, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2863, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "MISS_3_6_499_0", + "timepoint": 499, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 499, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; scalpel,patient,touching; head surgeon,herbal disk,removing; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,microscope eye,looking; curette,patient,touching; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1141_0", + "timepoint": 1141, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1141, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_510_0", + "timepoint": 510, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 510, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1945_0", + "timepoint": 1945, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1945, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; needle,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1709_0", + "timepoint": 1709, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1709, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,needle,inserting; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,tissue mark,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_597_0", + "timepoint": 597, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 597, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_161_0", + "timepoint": 161, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 161, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,cotton,looking; circulator,instruments,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_140_0", + "timepoint": 140, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 140, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2069_0", + "timepoint": 2069, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2069, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,disinfection; assistant,forceps,holding; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_467_0", + "timepoint": 467, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 467, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; curette,patient,touching; head surgeon,herbal disk,removing; head surgeon,curette,holding; head surgeon,forceps,holding; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_2493_0", + "timepoint": 2493, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2493, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,patient,manipulating; head surgeon,patient,manipulating; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1989_0", + "timepoint": 1989, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1989, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; ultrasound probe,patient,touching; needle,patient,touching; assistant,syringe,holding; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_245_0", + "timepoint": 245, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 245, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,ultrasound gel,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_310_0", + "timepoint": 310, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 310, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2720_0", + "timepoint": 2720, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2720, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; assistant,needle,holding; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; syringe,needle,touching; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_465_0", + "timepoint": 465, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 465, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_843_0", + "timepoint": 843, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 843, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2043_0", + "timepoint": 2043, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2043, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_938_0", + "timepoint": 938, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 938, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; patient,operating table,lyingon; forceps,cotton,touching; assistant,antiseptic,holding; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,patient,disinfection; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2684_0", + "timepoint": 2684, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2684, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; patient,operating table,lyingon; syringe,needle,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1385_0", + "timepoint": 1385, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1385, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1483_0", + "timepoint": 1483, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1483, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,microscope controller,manipulating; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,patient,anaesthetising; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_900_0", + "timepoint": 900, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 900, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2335_0", + "timepoint": 2335, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2335, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,syringe,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_355_0", + "timepoint": 355, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 355, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_152_0", + "timepoint": 152, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 152, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,patient,cutting; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,forceps,holding; assistant,instrument table,closeto; scalpel,patient,touching; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2586_0", + "timepoint": 2586, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2586, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,needle,holding; syringe,needle,touching; needle,patient,touching; assistant,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2521_0", + "timepoint": 2521, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2521, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2623_0", + "timepoint": 2623, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2623, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_1_1_2267_0", + "timepoint": 2267, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2267, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_15_0", + "timepoint": 15, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 15, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_939_0", + "timepoint": 939, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 939, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2210_0", + "timepoint": 2210, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2210, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; assistant,ultrasound screen,looking; syringe,needle,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3071_0", + "timepoint": 3071, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3071, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "MISS_1_1_1189_0", + "timepoint": 1189, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1189, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating room,entering; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1475_0", + "timepoint": 1475, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1475, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_1714_0", + "timepoint": 1714, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1714, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,looking; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1257_0", + "timepoint": 1257, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1257, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,antiseptic,holding; forceps,cotton,touching; assistant,forceps,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_139_0", + "timepoint": 139, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 139, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,forceps,holding; head surgeon,scalpel,holding; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,patient,cutting; scalpel,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2750_0", + "timepoint": 2750, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2750, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_260_0", + "timepoint": 260, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 260, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,patient,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_575_0", + "timepoint": 575, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 575, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1368_0", + "timepoint": 1368, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1368, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1925_0", + "timepoint": 1925, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1925, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; assistant,needle,applying; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_519_0", + "timepoint": 519, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 519, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,controlling; anesthetist,operating table,closeto; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_524_0", + "timepoint": 524, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 524, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_760_0", + "timepoint": 760, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 760, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instruments,disinfection; assistant,anesthetist,applying; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_781_0", + "timepoint": 781, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 781, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_937_0", + "timepoint": 937, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 937, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; circulator,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2046_0", + "timepoint": 2046, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2046, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,needle,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1144_0", + "timepoint": 1144, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1144, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,body marker,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2866_0", + "timepoint": 2866, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2866, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,ultrasound machine,looking; assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_888_0", + "timepoint": 888, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 888, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; forceps,cotton,touching; assistant,forceps,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,antiseptic,applying; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_4_5_1742_0", + "timepoint": 1742, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1742, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,needle,holding; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1260_0", + "timepoint": 1260, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1260, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_428_0", + "timepoint": 428, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 428, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,instrument table,closeto; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_839_0", + "timepoint": 839, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 839, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,antiseptic,holding; assistant,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2330_0", + "timepoint": 2330, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2330, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,holding; syringe,needle,touching; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_173_0", + "timepoint": 173, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 173, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; assistant,ultrasound gel,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_2979_0", + "timepoint": 2979, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2979, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,patient,scanning; assistant,operating table,closeto; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_445_0", + "timepoint": 445, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 445, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2953_0", + "timepoint": 2953, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2953, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,tissue paper,applying; tissue paper,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_991_0", + "timepoint": 991, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 991, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; patient,operating table,lyingon; forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,patient,disinfection; assistant,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_236_0", + "timepoint": 236, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 236, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,microscope eye,looking; anesthetist,health monitor,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,curette,holding; head surgeon,forceps,holding; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; curette,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_1116_0", + "timepoint": 1116, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1116, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; anesthetist,health monitor,looking; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2365_0", + "timepoint": 2365, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2365, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,syringe,holding; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2716_0", + "timepoint": 2716, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2716, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,bin,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_395_0", + "timepoint": 395, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 395, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; anesthetist,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "MISS_3_6_783_0", + "timepoint": 783, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 783, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; head surgeon,operating table,closeto; anesthetist,health monitor,looking; head surgeon,microscope eye,looking; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2577_0", + "timepoint": 2577, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2577, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,syringe,injecting; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_77_0", + "timepoint": 77, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 77, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,patient,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_380_0", + "timepoint": 380, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 380, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,ultrasound gel,applying; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_681_0", + "timepoint": 681, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 681, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,patient,manipulating; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2026_0", + "timepoint": 2026, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2026, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,syringe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; needle,patient,touching; assistant,needle,inserting; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2492_0", + "timepoint": 2492, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2492, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,syringe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1280_0", + "timepoint": 1280, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1280, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; anesthetist,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_880_0", + "timepoint": 880, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 880, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,body marker,applying; body marker,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_669_0", + "timepoint": 669, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 669, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,manipulating; head surgeon,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1044_0", + "timepoint": 1044, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1044, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,manipulating; anesthetist,microscope,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1574_0", + "timepoint": 1574, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1574, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,operating table,closeto; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2515_0", + "timepoint": 2515, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2515, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_744_0", + "timepoint": 744, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 744, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2385_0", + "timepoint": 2385, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2385, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; assistant,syringe,holding; ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_618_0", + "timepoint": 618, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 618, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; head surgeon,herbal disk,removing; head surgeon,scalpel,holding; anesthetist,health monitor,looking; head surgeon,operating table,closeto; curette,patient,touching; head surgeon,microscope eye,looking; scalpel,patient,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_282_0", + "timepoint": 282, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 282, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound gel,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1953_0", + "timepoint": 1953, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1953, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,syringe,holding; assistant,needle,inserting; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2741_0", + "timepoint": 2741, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2741, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2602_0", + "timepoint": 2602, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2602, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; syringe,needle,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,needle,holding; assistant,syringe,injecting; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1447_0", + "timepoint": 1447, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1447, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,patient,anaesthetising; anesthetist,operating table,closeto; head surgeon,microscope controller,manipulating; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1404_0", + "timepoint": 1404, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1404, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_980_0", + "timepoint": 980, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 980, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; assistant,forceps,holding; patient,operating table,lyingon; forceps,cotton,touching; cotton,patient,touching; head surgeon,operating table,closeto; assistant,antiseptic,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2344_0", + "timepoint": 2344, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2344, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; syringe,needle,touching; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; needle,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2660_0", + "timepoint": 2660, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2660, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,tissue paper,holding; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_516_0", + "timepoint": 516, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 516, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating room,entering; circulator,assistant,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3186_0", + "timepoint": 3186, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3186, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,tissue paper,applying; assistant,operating table,closeto; tissue paper,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1544_0", + "timepoint": 1544, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1544, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_756_0", + "timepoint": 756, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 756, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_550_0", + "timepoint": 550, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 550, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_456_0", + "timepoint": 456, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 456, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_254_0", + "timepoint": 254, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 254, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,looking; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_2230_0", + "timepoint": 2230, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2230, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,needle,holding; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_431_0", + "timepoint": 431, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 431, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; assistant,instrument table,closeto; head surgeon,forceps,holding; anesthetist,anesthesia equipment,controlling; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; patient,operating table,lyingon; curette,patient,touching; anesthetist,operating table,closeto; assistant,scalpel,holding; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2816_0", + "timepoint": 2816, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2816, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,ultrasound machine,looking; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_284_0", + "timepoint": 284, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 284, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1045_0", + "timepoint": 1045, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1045, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,health monitor,looking; assistant,patient,aspirating; anesthetist,operating table,closeto; assistant,forceps,holding; head surgeon,microscope eye,looking; assistant,operating table,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_2374_0", + "timepoint": 2374, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2374, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1533_0", + "timepoint": 1533, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1533, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2451_0", + "timepoint": 2451, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2451, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; syringe,needle,touching; head surgeon,needle,holding; assistant,syringe,injecting; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1963_0", + "timepoint": 1963, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1963, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,syringe,injecting; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2868_0", + "timepoint": 2868, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2868, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; assistant,instrument table,closeto; assistant,needle,holding; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1702_0", + "timepoint": 1702, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1702, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,inserting; assistant,tissue mark,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_255_0", + "timepoint": 255, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 255, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,patient,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_748_0", + "timepoint": 748, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 748, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_594_0", + "timepoint": 594, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 594, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1615_0", + "timepoint": 1615, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1615, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_620_0", + "timepoint": 620, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 620, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,microscope screen,looking; head surgeon,operating table,closeto; anesthetist,health monitor,looking; assistant,operating table,closeto; head surgeon,microscope controller,controlling; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,microscope screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1263_0", + "timepoint": 1263, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1263, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2382_0", + "timepoint": 2382, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2382, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_389_0", + "timepoint": 389, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 389, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "MISS_3_6_510_0", + "timepoint": 510, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 510, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,scalpel,holding; scalpel,patient,touching; patient,operating table,lyingon; head surgeon,herbal disk,removing; head surgeon,microscope eye,looking; curette,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_201_0", + "timepoint": 201, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 201, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1875_0", + "timepoint": 1875, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1875, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_3_4_252_0", + "timepoint": 252, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 252, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound gel,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2453_0", + "timepoint": 2453, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2453, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1686_0", + "timepoint": 1686, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1686, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_3042_0", + "timepoint": 3042, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3042, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,tissue paper,applying; tissue paper,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2444_0", + "timepoint": 2444, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2444, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound machine,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_301_0", + "timepoint": 301, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 301, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_598_0", + "timepoint": 598, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 598, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1185_0", + "timepoint": 1185, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1185, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; circulator,ultrasound machine,manipulating; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,operating table,closeto; circulator,ultrasound machine,closeto; patient,operating table,lyingon; assistant,antiseptic,holding " + } + ] + }, + { + "id": "MISS_1_1_1673_0", + "timepoint": 1673, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1673, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1218_0", + "timepoint": 1218, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1218, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,body marker,holding; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_980_0", + "timepoint": 980, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 980, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1675_0", + "timepoint": 1675, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1675, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; assistant,needle,inserting; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,tissue mark,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_3324_0", + "timepoint": 3324, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3324, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,bin,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_152_0", + "timepoint": 152, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 152, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,patient,manipulating; head surgeon,ultrasound machine,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1979_0", + "timepoint": 1979, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1979, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,patient,disinfection; assistant,instrument table,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1059_0", + "timepoint": 1059, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1059, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_62_0", + "timepoint": 62, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 62, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2528_0", + "timepoint": 2528, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2528, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching; head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_928_0", + "timepoint": 928, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 928, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2254_0", + "timepoint": 2254, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2254, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,patient,disinfection; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2379_0", + "timepoint": 2379, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2379, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1245_0", + "timepoint": 1245, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1245, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,applying; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_880_0", + "timepoint": 880, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 880, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_2_2_632_0", + "timepoint": 632, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 632, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,controlling; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,microscope screen,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,microscope screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_326_0", + "timepoint": 326, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 326, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,looking; circulator,patient,closeto; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_2708_0", + "timepoint": 2708, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2708, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,tissue paper,dropping; head surgeon,operating table,closeto; assistant,bin,closeto; assistant,bin,looking " + } + ] + }, + { + "id": "MISS_1_1_703_0", + "timepoint": 703, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 703, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1313_0", + "timepoint": 1313, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1313, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1374_0", + "timepoint": 1374, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1374, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_795_0", + "timepoint": 795, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 795, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_840_0", + "timepoint": 840, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 840, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1791_0", + "timepoint": 1791, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1791, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2484_0", + "timepoint": 2484, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2484, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_720_0", + "timepoint": 720, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 720, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_717_0", + "timepoint": 717, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 717, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_903_0", + "timepoint": 903, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 903, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_211_0", + "timepoint": 211, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 211, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; anesthetist,operating table,closeto; head surgeon,scalpel,holding; scalpel,patient,touching; patient,operating table,lyingon; forceps,cotton,touching; assistant,forceps,holding; assistant,operating table,closeto; head surgeon,patient,cutting; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2300_0", + "timepoint": 2300, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2300, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_542_0", + "timepoint": 542, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 542, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2569_0", + "timepoint": 2569, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2569, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_840_0", + "timepoint": 840, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 840, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1020_0", + "timepoint": 1020, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1020, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,body marker,holding; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2666_0", + "timepoint": 2666, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2666, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,needle,holding; patient,operating table,lyingon; assistant,needle,holding; assistant,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_818_0", + "timepoint": 818, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 818, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2187_0", + "timepoint": 2187, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2187, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; needle,patient,touching; syringe,needle,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_1_1_17_0", + "timepoint": 17, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 17, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2273_0", + "timepoint": 2273, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2273, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,needle,holding; ultrasound probe,patient,touching; needle,patient,touching; assistant,syringe,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2251_0", + "timepoint": 2251, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2251, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; needle,patient,touching; patient,operating table,lyingon; syringe,needle,touching; assistant,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3239_0", + "timepoint": 3239, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3239, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; tissue paper,patient,touching; assistant,tissue paper,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2655_0", + "timepoint": 2655, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2655, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1440_0", + "timepoint": 1440, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1440, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1759_0", + "timepoint": 1759, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1759, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,needle,inserting; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,tissue mark,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1942_0", + "timepoint": 1942, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1942, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,syringe,injecting; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_966_0", + "timepoint": 966, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 966, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,antiseptic,holding; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; cotton,patient,touching; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_3_4_1348_0", + "timepoint": 1348, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1348, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_508_0", + "timepoint": 508, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 508, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_432_0", + "timepoint": 432, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 432, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,scalpel,holding; assistant,instrument table,closeto; head surgeon,microscope eye,looking; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,forceps,holding; anesthetist,operating table,closeto; curette,patient,touching; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "Ultrasound_5_6_1564_0", + "timepoint": 1564, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1564, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1095_0", + "timepoint": 1095, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1095, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,inserting; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2517_0", + "timepoint": 2517, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2517, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; head surgeon,patient,manipulating; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_747_0", + "timepoint": 747, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 747, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instruments,disinfection; assistant,instrument table,closeto; anesthetist,health monitor,looking; assistant,anesthetist,applying; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1358_0", + "timepoint": 1358, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1358, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,removing; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_211_0", + "timepoint": 211, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 211, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,ultrasound gel,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2494_0", + "timepoint": 2494, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2494, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,tissue paper,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_348_0", + "timepoint": 348, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 348, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,herbal disk,removing; head surgeon,curette,holding; head surgeon,operating table,closeto; curette,patient,touching; forceps,patient,touching; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1093_0", + "timepoint": 1093, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1093, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; head surgeon,operating table,closeto; assistant,scissors,manipulating; assistant,patient,disinfection; patient,operating table,lyingon; cotton,patient,touching; scissors,cotton,touching; assistant,operating table,closeto; circulator,ultrasound machine,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_844_0", + "timepoint": 844, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 844, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2429_0", + "timepoint": 2429, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2429, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,needle,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,syringe,injecting; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_205_0", + "timepoint": 205, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 205, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_2695_0", + "timepoint": 2695, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2695, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2065_0", + "timepoint": 2065, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2065, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; syringe,needle,touching; assistant,operating table,closeto; head surgeon,needle,holding; assistant,syringe,injecting; head surgeon,vertebrae,scanning; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_383_0", + "timepoint": 383, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 383, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; anesthetist,operating table,manipulating; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_3062_0", + "timepoint": 3062, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3062, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,tissue paper,applying; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_992_0", + "timepoint": 992, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 992, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; forceps,cotton,touching; assistant,forceps,holding; assistant,patient,disinfection; cotton,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2528_0", + "timepoint": 2528, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2528, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,tissue paper,holding; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_266_0", + "timepoint": 266, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 266, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,patient,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_545_0", + "timepoint": 545, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 545, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2381_0", + "timepoint": 2381, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2381, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_559_0", + "timepoint": 559, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 559, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_324_0", + "timepoint": 324, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 324, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound gel,applying; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2945_0", + "timepoint": 2945, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2945, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; patient,operating table,lyingon; head surgeon,tissue paper,applying; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_805_0", + "timepoint": 805, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 805, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2925_0", + "timepoint": 2925, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2925, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,tissue paper,holding; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1967_0", + "timepoint": 1967, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1967, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,syringe,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_376_0", + "timepoint": 376, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 376, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "MISS_1_1_1603_0", + "timepoint": 1603, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1603, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_424_0", + "timepoint": 424, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 424, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,touching; circulator,operating table,closeto; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1340_0", + "timepoint": 1340, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1340, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_354_0", + "timepoint": 354, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 354, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_479_0", + "timepoint": 479, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 479, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,health monitor,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3067_0", + "timepoint": 3067, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3067, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,tissue paper,holding; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1036_0", + "timepoint": 1036, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1036, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1936_0", + "timepoint": 1936, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1936, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound probe,looking; assistant,syringe,holding; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1390_0", + "timepoint": 1390, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1390, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,needle,removing " + } + ] + }, + { + "id": "Ultrasound_2_2_61_0", + "timepoint": 61, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 61, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_451_0", + "timepoint": 451, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 451, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2475_0", + "timepoint": 2475, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2475, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; syringe,needle,touching; needle,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_1716_0", + "timepoint": 1716, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1716, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_717_0", + "timepoint": 717, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 717, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_863_0", + "timepoint": 863, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 863, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,microscope,manipulating; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "MISS_3_6_246_0", + "timepoint": 246, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 246, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; anesthetist,health monitor,looking; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; assistant,operating table,closeto; head surgeon,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; forceps,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1529_0", + "timepoint": 1529, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1529, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_227_0", + "timepoint": 227, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 227, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,instrument table,looking; assistant,ultrasound gel,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_763_0", + "timepoint": 763, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 763, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instruments,disinfection; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; assistant,anesthetist,applying; assistant,instrument table,closeto; anesthetist,operating table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "MISS_2_2_375_0", + "timepoint": 375, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 375, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; cotton,patient,touching; assistant,patient,aspirating; head surgeon,scalpel,holding; assistant,operating table,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_745_0", + "timepoint": 745, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 745, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_884_0", + "timepoint": 884, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 884, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_3049_0", + "timepoint": 3049, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3049, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2666_0", + "timepoint": 2666, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2666, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,syringe,injecting; assistant,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; syringe,needle,touching; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1167_0", + "timepoint": 1167, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1167, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_2021_0", + "timepoint": 2021, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2021, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,patient,disinfection; patient,operating table,lyingon; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_3_6_851_0", + "timepoint": 851, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 851, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_34_0", + "timepoint": 34, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 34, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,patient,manipulating; assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1692_0", + "timepoint": 1692, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1692, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1113_0", + "timepoint": 1113, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1113, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_3_6_1197_0", + "timepoint": 1197, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1197, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1212_0", + "timepoint": 1212, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1212, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2625_0", + "timepoint": 2625, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2625, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; needle,patient,touching; head surgeon,ultrasound probe,controlling; syringe,needle,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,holding; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_1_1_254_0", + "timepoint": 254, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 254, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; anesthetist,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_796_0", + "timepoint": 796, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 796, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_798_0", + "timepoint": 798, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 798, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,microscope screen,looking; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_394_0", + "timepoint": 394, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 394, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,forceps,holding; anesthetist,operating table,closeto; assistant,forceps,holding; head surgeon,microscope eye,looking; curette,patient,touching; head surgeon,operating table,closeto; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; head surgeon,herbal disk,removing; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2260_0", + "timepoint": 2260, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2260, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; forceps,cotton,touching; cotton,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,forceps,holding; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_4_5_2211_0", + "timepoint": 2211, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2211, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; syringe,needle,touching; head surgeon,needle,holding; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_897_0", + "timepoint": 897, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 897, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; head surgeon,microscope eye,looking; forceps,cotton,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_932_0", + "timepoint": 932, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 932, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_38_0", + "timepoint": 38, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 38, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_1_1_1695_0", + "timepoint": 1695, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1695, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_197_0", + "timepoint": 197, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 197, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_1_1_2282_0", + "timepoint": 2282, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2282, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2723_0", + "timepoint": 2723, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2723, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; syringe,needle,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,needle,holding; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_424_0", + "timepoint": 424, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 424, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; assistant,instrument table,closeto; head surgeon,ultrasound probe,holding; assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_875_0", + "timepoint": 875, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 875, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " body marker,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,ultrasound probe,looking; assistant,body marker,applying; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1150_0", + "timepoint": 1150, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1150, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,body marker,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1595_0", + "timepoint": 1595, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1595, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_376_0", + "timepoint": 376, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 376, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,aspirating; forceps,cotton,touching; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,scalpel,holding; cotton,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_2467_0", + "timepoint": 2467, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2467, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,patient,manipulating; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1456_0", + "timepoint": 1456, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1456, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_969_0", + "timepoint": 969, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 969, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2445_0", + "timepoint": 2445, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2445, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "MISS_1_1_2206_0", + "timepoint": 2206, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2206, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; forceps,cotton,touching; assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_373_0", + "timepoint": 373, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 373, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1839_0", + "timepoint": 1839, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1839, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,vertebrae,scanning; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_497_0", + "timepoint": 497, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 497, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; anesthetist,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_1656_0", + "timepoint": 1656, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1656, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,needle,inserting; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1842_0", + "timepoint": 1842, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1842, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,needle,holding; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1904_0", + "timepoint": 1904, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1904, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2517_0", + "timepoint": 2517, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2517, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; syringe,needle,touching; patient,operating table,lyingon; assistant,needle,holding; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_503_0", + "timepoint": 503, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 503, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2259_0", + "timepoint": 2259, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2259, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,ultrasound screen,looking; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_289_0", + "timepoint": 289, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 289, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating room,entering; anesthetist,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_887_0", + "timepoint": 887, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 887, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; assistant,instrument table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_484_0", + "timepoint": 484, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 484, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2996_0", + "timepoint": 2996, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2996, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_1591_0", + "timepoint": 1591, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1591, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,needle,inserting; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1583_0", + "timepoint": 1583, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1583, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,needle,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1233_0", + "timepoint": 1233, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1233, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; assistant,operating table,closeto; assistant,body marker,applying; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_996_0", + "timepoint": 996, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 996, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_2_2_746_0", + "timepoint": 746, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 746, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,microscope screen,looking " + } + ] + }, + { + "id": "MISS_1_1_1242_0", + "timepoint": 1242, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1242, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,patient,manipulating; anesthetist,anesthesia equipment,holding; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_266_0", + "timepoint": 266, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 266, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_278_0", + "timepoint": 278, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 278, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; head surgeon,forceps,holding; anesthetist,health monitor,looking; forceps,patient,touching; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,curette,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_994_0", + "timepoint": 994, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 994, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; cotton,patient,touching; patient,operating table,lyingon; assistant,forceps,holding; forceps,cotton,touching; assistant,patient,disinfection; assistant,antiseptic,holding " + } + ] + }, + { + "id": "MISS_1_1_2172_0", + "timepoint": 2172, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2172, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,operating table,closeto; assistant,patient,disinfection; patient,operating table,lyingon; cotton,patient,touching; assistant,forceps,holding; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1796_0", + "timepoint": 1796, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1796, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,syringe,holding; assistant,needle,holding; assistant,instrument table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_2330_0", + "timepoint": 2330, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2330, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3149_0", + "timepoint": 3149, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3149, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; tissue paper,patient,touching; assistant,tissue paper,applying; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1442_0", + "timepoint": 1442, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1442, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2395_0", + "timepoint": 2395, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2395, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_345_0", + "timepoint": 345, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 345, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,microscope eye,looking; forceps,patient,touching; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; curette,patient,touching; head surgeon,curette,holding; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_597_0", + "timepoint": 597, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 597, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1337_0", + "timepoint": 1337, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1337, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3041_0", + "timepoint": 3041, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3041, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; assistant,operating table,closeto; patient,operating table,lyingon; tissue paper,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_46_0", + "timepoint": 46, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 46, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,forceps,looking; circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2705_0", + "timepoint": 2705, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2705, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; syringe,needle,touching; assistant,syringe,injecting; needle,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_448_0", + "timepoint": 448, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 448, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_561_0", + "timepoint": 561, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 561, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1689_0", + "timepoint": 1689, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1689, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_597_0", + "timepoint": 597, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 597, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,health monitor,looking; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope controller,controlling; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2467_0", + "timepoint": 2467, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2467, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_1225_0", + "timepoint": 1225, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1225, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,patient,manipulating; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1217_0", + "timepoint": 1217, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1217, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; circulator,ultrasound machine,manipulating; assistant,forceps,holding; circulator,ultrasound machine,closeto " + } + ] + }, + { + "id": "MISS_3_6_896_0", + "timepoint": 896, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 896, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; forceps,cotton,touching; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_480_0", + "timepoint": 480, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 480, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,curette,holding; head surgeon,herbal disk,removing; curette,patient,touching; assistant,operating table,closeto; head surgeon,scalpel,holding; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; scalpel,patient,touching; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_533_0", + "timepoint": 533, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 533, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2676_0", + "timepoint": 2676, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2676, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_458_0", + "timepoint": 458, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 458, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; patient,operating table,lyingon; assistant,operating table,closeto; curette,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; head surgeon,curette,holding; head surgeon,herbal disk,removing; head surgeon,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_618_0", + "timepoint": 618, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 618, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_399_0", + "timepoint": 399, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 399, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding; head surgeon,operating table,closeto; assistant,instrument table,looking; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2124_0", + "timepoint": 2124, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2124, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; ultrasound probe,patient,touching; assistant,needle,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2525_0", + "timepoint": 2525, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2525, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,patient,manipulating; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_432_0", + "timepoint": 432, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 432, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound gel,holding; assistant,instrument table,looking; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1251_0", + "timepoint": 1251, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1251, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2120_0", + "timepoint": 2120, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2120, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; patient,operating table,lyingon; assistant,patient,disinfection; forceps,cotton,touching; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_200_0", + "timepoint": 200, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 200, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; assistant,ultrasound gel,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_323_0", + "timepoint": 323, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 323, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,herbal disk,removing; assistant,operating table,closeto; curette,patient,touching; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; head surgeon,forceps,holding; forceps,patient,touching; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "MISS_3_6_460_0", + "timepoint": 460, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 460, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; curette,patient,touching; head surgeon,herbal disk,removing; head surgeon,scalpel,holding; head surgeon,curette,holding; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,forceps,holding; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_565_0", + "timepoint": 565, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 565, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1397_0", + "timepoint": 1397, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1397, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,removing; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_3058_0", + "timepoint": 3058, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3058, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; head surgeon,operating table,closeto; head surgeon,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1007_0", + "timepoint": 1007, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1007, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,body marker,holding; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_394_0", + "timepoint": 394, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 394, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1004_0", + "timepoint": 1004, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1004, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_2094_0", + "timepoint": 2094, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2094, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,patient,disinfection; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; cotton,patient,touching; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_602_0", + "timepoint": 602, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 602, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1451_0", + "timepoint": 1451, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1451, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1761_0", + "timepoint": 1761, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1761, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,needle,inserting; assistant,tissue mark,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1290_0", + "timepoint": 1290, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1290, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,body marker,applying; body marker,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_123_0", + "timepoint": 123, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 123, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,scalpel,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2288_0", + "timepoint": 2288, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2288, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,needle,holding; needle,patient,touching; assistant,needle,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,injecting; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_3_6_989_0", + "timepoint": 989, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 989, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling; assistant,patient,aspirating; cotton,patient,touching; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2331_0", + "timepoint": 2331, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2331, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; needle,patient,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; syringe,needle,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_739_0", + "timepoint": 739, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 739, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,instruments,disinfection; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; anesthetist,health monitor,looking; assistant,anesthetist,applying " + } + ] + }, + { + "id": "MISS_3_6_579_0", + "timepoint": 579, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 579, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,herbal disk,removing; head surgeon,curette,holding; curette,patient,touching; head surgeon,scalpel,holding; scalpel,patient,touching; head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_154_0", + "timepoint": 154, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 154, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; assistant,operating table,closeto; assistant,forceps,holding; head surgeon,curette,holding; head surgeon,herbal disk,removing; patient,operating table,lyingon; curette,patient,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1038_0", + "timepoint": 1038, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1038, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1_0", + "timepoint": 1, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,microscope controller,controlling; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_689_0", + "timepoint": 689, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 689, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,patient,manipulating; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking " + } + ] + }, + { + "id": "MISS_1_1_1184_0", + "timepoint": 1184, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1184, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating room,entering; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1219_0", + "timepoint": 1219, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1219, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,body marker,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3120_0", + "timepoint": 3120, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3120, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1439_0", + "timepoint": 1439, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1439, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,needle,removing; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1052_0", + "timepoint": 1052, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1052, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,body marker,holding; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_59_0", + "timepoint": 59, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 59, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound machine,manipulating; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1546_0", + "timepoint": 1546, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1546, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,patient,anaesthetising; head surgeon,microscope controller,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_312_0", + "timepoint": 312, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 312, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; anesthetist,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_644_0", + "timepoint": 644, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 644, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1039_0", + "timepoint": 1039, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1039, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; forceps,cotton,touching; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,health monitor,looking; cotton,patient,touching; assistant,operating table,closeto; assistant,patient,aspirating " + } + ] + }, + { + "id": "Ultrasound_3_4_1067_0", + "timepoint": 1067, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1067, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_996_0", + "timepoint": 996, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 996, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1238_0", + "timepoint": 1238, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1238, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,antiseptic,holding; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_892_0", + "timepoint": 892, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 892, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,closeto; forceps,cotton,touching; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_1_1_1929_0", + "timepoint": 1929, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1929, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,patient,disinfection; assistant,antiseptic,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_792_0", + "timepoint": 792, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 792, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2952_0", + "timepoint": 2952, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2952, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; patient,operating table,lyingon; tissue paper,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_712_0", + "timepoint": 712, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 712, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,microscope screen,looking; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3069_0", + "timepoint": 3069, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3069, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,tissue paper,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2614_0", + "timepoint": 2614, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2614, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,needle,holding; syringe,needle,touching; patient,operating table,lyingon; head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_4_5_333_0", + "timepoint": 333, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 333, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound gel,applying; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1301_0", + "timepoint": 1301, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1301, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; assistant,needle,inserting; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_1248_0", + "timepoint": 1248, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1248, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,patient,manipulating; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "MISS_1_1_1597_0", + "timepoint": 1597, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1597, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1096_0", + "timepoint": 1096, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1096, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_437_0", + "timepoint": 437, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 437, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1176_0", + "timepoint": 1176, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1176, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; patient,operating table,lyingon; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_365_0", + "timepoint": 365, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 365, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_554_0", + "timepoint": 554, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 554, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating room,entering; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1899_0", + "timepoint": 1899, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1899, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,syringe,holding; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1724_0", + "timepoint": 1724, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1724, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; assistant,tissue mark,looking; head surgeon,operating table,closeto; assistant,needle,inserting; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_366_0", + "timepoint": 366, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 366, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2573_0", + "timepoint": 2573, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2573, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; syringe,needle,touching; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,needle,holding; assistant,needle,holding; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3018_0", + "timepoint": 3018, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3018, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,tissue paper,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1315_0", + "timepoint": 1315, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1315, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; body marker,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,body marker,applying; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_552_0", + "timepoint": 552, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 552, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2261_0", + "timepoint": 2261, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2261, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; cotton,patient,touching; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,patient,disinfection; assistant,operating table,closeto; forceps,cotton,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2284_0", + "timepoint": 2284, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2284, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound probe,controlling; syringe,needle,touching; ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_2_2_615_0", + "timepoint": 615, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 615, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,controlling; anesthetist,anesthesia equipment,controlling; assistant,microscope screen,looking; anesthetist,health monitor,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2431_0", + "timepoint": 2431, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2431, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; head surgeon,ultrasound probe,holding; head surgeon,ultrasound machine,looking; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1020_0", + "timepoint": 1020, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1020, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; patient,operating table,lyingon; assistant,patient,disinfection; assistant,forceps,holding; forceps,cotton,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2222_0", + "timepoint": 2222, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2222, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,needle,holding; syringe,needle,touching; ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_1_1_2245_0", + "timepoint": 2245, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2245, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; cotton,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1000_0", + "timepoint": 1000, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1000, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1908_0", + "timepoint": 1908, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1908, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,patient,disinfection; assistant,operating table,closeto; assistant,antiseptic,applying; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2481_0", + "timepoint": 2481, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2481, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon; head surgeon,needle,holding; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2706_0", + "timepoint": 2706, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2706, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,bin,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,bin,closeto; assistant,tissue paper,dropping " + } + ] + }, + { + "id": "MISS_3_6_574_0", + "timepoint": 574, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 574, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,curette,holding; scalpel,patient,touching; assistant,operating table,closeto; head surgeon,herbal disk,removing; curette,patient,touching; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1152_0", + "timepoint": 1152, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1152, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_347_0", + "timepoint": 347, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 347, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,looking; patient,operating table,closeto; circulator,operating table,closeto; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1573_0", + "timepoint": 1573, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1573, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_587_0", + "timepoint": 587, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 587, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2354_0", + "timepoint": 2354, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2354, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,removing " + } + ] + }, + { + "id": "Ultrasound_3_4_2384_0", + "timepoint": 2384, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2384, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; assistant,syringe,injecting; assistant,needle,holding; assistant,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2217_0", + "timepoint": 2217, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2217, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; needle,patient,touching; head surgeon,needle,holding; assistant,needle,holding; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_1_1_1619_0", + "timepoint": 1619, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1619, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_785_0", + "timepoint": 785, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 785, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; circulator,microscope,manipulating; circulator,microscope,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1122_0", + "timepoint": 1122, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1122, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1785_0", + "timepoint": 1785, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1785, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,tissue mark,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,needle,inserting; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2052_0", + "timepoint": 2052, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2052, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; syringe,needle,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1513_0", + "timepoint": 1513, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1513, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1552_0", + "timepoint": 1552, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1552, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,needle,inserting; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1237_0", + "timepoint": 1237, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1237, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching; assistant,forceps,holding; assistant,operating table,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1412_0", + "timepoint": 1412, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1412, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_286_0", + "timepoint": 286, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 286, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,scalpel,holding; assistant,forceps,holding; anesthetist,health monitor,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_1389_0", + "timepoint": 1389, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1389, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,holding; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,patient,anaesthetising; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1904_0", + "timepoint": 1904, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1904, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,disinfection; anesthetist,operating table,closeto; assistant,antiseptic,applying; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1107_0", + "timepoint": 1107, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1107, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; assistant,forceps,holding; assistant,cotton,holding; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2244_0", + "timepoint": 2244, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2244, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; head surgeon,operating table,closeto; cotton,patient,touching; assistant,forceps,holding; forceps,cotton,touching; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_44_0", + "timepoint": 44, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 44, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; anesthetist,operating table,closeto; head surgeon,patient,manipulating; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2934_0", + "timepoint": 2934, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2934, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; tissue paper,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_995_0", + "timepoint": 995, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 995, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_216_0", + "timepoint": 216, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 216, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,scalpel,holding; head surgeon,patient,cutting; scalpel,patient,touching; anesthetist,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_126_0", + "timepoint": 126, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 126, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3275_0", + "timepoint": 3275, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3275, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,tissue paper,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2946_0", + "timepoint": 2946, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2946, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying; tissue paper,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_402_0", + "timepoint": 402, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 402, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2452_0", + "timepoint": 2452, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2452, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; syringe,needle,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_275_0", + "timepoint": 275, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 275, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; head surgeon,operating table,closeto; head surgeon,scalpel,holding; forceps,cotton,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1593_0", + "timepoint": 1593, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1593, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1262_0", + "timepoint": 1262, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1262, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_5_6_2859_0", + "timepoint": 2859, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2859, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,syringe,removing; assistant,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; syringe,needle,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_95_0", + "timepoint": 95, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 95, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,antiseptic,looking; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_1_1_1679_0", + "timepoint": 1679, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1679, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2966_0", + "timepoint": 2966, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2966, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; needle,patient,touching; head surgeon,patient,scanning; head surgeon,operating table,closeto; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_17_0", + "timepoint": 17, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 17, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_719_0", + "timepoint": 719, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 719, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,microscope screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_195_0", + "timepoint": 195, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 195, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; head surgeon,herbal disk,removing; assistant,operating table,closeto; curette,patient,touching; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,forceps,holding " + } + ] + }, + { + "id": "MISS_1_1_2072_0", + "timepoint": 2072, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2072, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,forceps,holding; assistant,patient,disinfection; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2179_0", + "timepoint": 2179, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2179, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,syringe,injecting; assistant,needle,holding; head surgeon,ultrasound screen,looking; needle,patient,touching; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_140_0", + "timepoint": 140, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 140, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,applying; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2869_0", + "timepoint": 2869, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2869, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound machine,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,needle,holding; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_183_0", + "timepoint": 183, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 183, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_903_0", + "timepoint": 903, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 903, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,microscope,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2108_0", + "timepoint": 2108, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2108, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; assistant,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1785_0", + "timepoint": 1785, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1785, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,syringe,holding; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_3135_0", + "timepoint": 3135, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3135, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; tissue paper,patient,touching; assistant,tissue paper,applying; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1985_0", + "timepoint": 1985, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1985, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1097_0", + "timepoint": 1097, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1097, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1778_0", + "timepoint": 1778, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1778, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,syringe,holding; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_407_0", + "timepoint": 407, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 407, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2699_0", + "timepoint": 2699, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2699, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,bin,closeto; patient,operating table,lyingon; assistant,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1329_0", + "timepoint": 1329, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1329, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,antiseptic,holding; patient,operating table,lyingon; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_2359_0", + "timepoint": 2359, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2359, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2160_0", + "timepoint": 2160, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2160, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,syringe,injecting; assistant,ultrasound screen,looking; needle,patient,touching; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_728_0", + "timepoint": 728, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 728, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1560_0", + "timepoint": 1560, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1560, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,patient,anaesthetising; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1200_0", + "timepoint": 1200, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1200, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1431_0", + "timepoint": 1431, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1431, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,patient,anaesthetising; head surgeon,microscope controller,manipulating; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_562_0", + "timepoint": 562, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 562, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; curette,patient,touching; head surgeon,herbal disk,removing; head surgeon,curette,holding; head surgeon,scalpel,holding; patient,operating table,lyingon; scalpel,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2231_0", + "timepoint": 2231, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2231, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; head surgeon,needle,holding; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_1166_0", + "timepoint": 1166, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1166, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1830_0", + "timepoint": 1830, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1830, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,inserting; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1438_0", + "timepoint": 1438, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1438, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,microscope controller,manipulating; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; anesthetist,patient,anaesthetising; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_31_0", + "timepoint": 31, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 31, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking " + } + ] + }, + { + "id": "MISS_2_2_529_0", + "timepoint": 529, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 529, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,health monitor,looking; circulator,instrument table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,microscope controller,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1486_0", + "timepoint": 1486, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1486, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,holding; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1662_0", + "timepoint": 1662, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1662, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,inserting; assistant,tissue mark,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2683_0", + "timepoint": 2683, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2683, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; syringe,needle,touching; ultrasound probe,patient,touching; needle,patient,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_283_0", + "timepoint": 283, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 283, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_399_0", + "timepoint": 399, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 399, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; anesthetist,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_65_0", + "timepoint": 65, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 65, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2381_0", + "timepoint": 2381, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2381, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,removing; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2237_0", + "timepoint": 2237, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2237, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2857_0", + "timepoint": 2857, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2857, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,removing; assistant,needle,holding; syringe,needle,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_628_0", + "timepoint": 628, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 628, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_2_2_141_0", + "timepoint": 141, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 141, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,patient,cutting; head surgeon,scalpel,holding; assistant,instrument table,closeto; scalpel,patient,touching; assistant,forceps,holding; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2806_0", + "timepoint": 2806, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2806, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; assistant,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2011_0", + "timepoint": 2011, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2011, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; syringe,needle,touching; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_1_1_443_0", + "timepoint": 443, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 443, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1177_0", + "timepoint": 1177, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1177, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating room,entering; head surgeon,operating room,entering; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1320_0", + "timepoint": 1320, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1320, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,inserting; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_1524_0", + "timepoint": 1524, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1524, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,patient,anaesthetising; head surgeon,microscope controller,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2076_0", + "timepoint": 2076, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2076, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto; syringe,needle,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1334_0", + "timepoint": 1334, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1334, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1999_0", + "timepoint": 1999, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1999, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,operating table,closeto; syringe,needle,touching; assistant,syringe,injecting; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_649_0", + "timepoint": 649, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 649, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1033_0", + "timepoint": 1033, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1033, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; assistant,needle,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2305_0", + "timepoint": 2305, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2305, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; needle,patient,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_553_0", + "timepoint": 553, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 553, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2046_0", + "timepoint": 2046, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2046, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; syringe,needle,touching; assistant,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2532_0", + "timepoint": 2532, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2532, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; syringe,needle,touching; assistant,syringe,injecting; patient,operating table,lyingon; assistant,needle,holding; needle,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_275_0", + "timepoint": 275, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 275, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1071_0", + "timepoint": 1071, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1071, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2481_0", + "timepoint": 2481, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2481, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,manipulating; head surgeon,patient,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_184_0", + "timepoint": 184, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 184, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1974_0", + "timepoint": 1974, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1974, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; assistant,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2629_0", + "timepoint": 2629, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2629, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2233_0", + "timepoint": 2233, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2233, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1498_0", + "timepoint": 1498, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1498, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_120_0", + "timepoint": 120, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 120, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling; circulator,syringe,looking " + } + ] + }, + { + "id": "MISS_3_6_330_0", + "timepoint": 330, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 330, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,health monitor,looking; forceps,patient,touching; curette,patient,touching; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,forceps,holding; anesthetist,operating table,closeto; head surgeon,curette,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_219_0", + "timepoint": 219, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 219, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2663_0", + "timepoint": 2663, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2663, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; patient,operating table,lyingon; assistant,syringe,injecting; needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2177_0", + "timepoint": 2177, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2177, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,operating table,closeto; needle,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; syringe,needle,touching; patient,operating table,lyingon; head surgeon,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2312_0", + "timepoint": 2312, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2312, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,needle,holding; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; needle,patient,touching; assistant,syringe,injecting; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_484_0", + "timepoint": 484, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 484, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1208_0", + "timepoint": 1208, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1208, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,inserting; head surgeon,vertebrae,scanning; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_26_0", + "timepoint": 26, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 26, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound machine,looking; assistant,instrument table,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1261_0", + "timepoint": 1261, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1261, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding; head surgeon,operating table,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1370_0", + "timepoint": 1370, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1370, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,removing; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1084_0", + "timepoint": 1084, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1084, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,body marker,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2441_0", + "timepoint": 2441, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2441, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,instrument table,looking; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1788_0", + "timepoint": 1788, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1788, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; assistant,needle,inserting; assistant,tissue mark,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2045_0", + "timepoint": 2045, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2045, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,syringe,holding; assistant,ultrasound probe,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_5_6_699_0", + "timepoint": 699, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 699, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_877_0", + "timepoint": 877, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 877, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1963_0", + "timepoint": 1963, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1963, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_5_6_2222_0", + "timepoint": 2222, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2222, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; syringe,needle,touching; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1429_0", + "timepoint": 1429, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1429, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_759_0", + "timepoint": 759, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 759, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope screen,looking; assistant,operating table,closeto; assistant,microscope screen,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_970_0", + "timepoint": 970, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 970, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,antiseptic,holding; forceps,cotton,touching; assistant,forceps,holding; patient,operating table,lyingon; cotton,patient,touching; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_4_5_1267_0", + "timepoint": 1267, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1267, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; patient,operating table,lyingon; assistant,needle,inserting; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1163_0", + "timepoint": 1163, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1163, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_232_0", + "timepoint": 232, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 232, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_935_0", + "timepoint": 935, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 935, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3065_0", + "timepoint": 3065, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3065, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,tissue paper,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_197_0", + "timepoint": 197, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 197, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; head surgeon,curette,holding; assistant,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; head surgeon,forceps,holding; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; curette,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_963_0", + "timepoint": 963, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 963, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,body marker,holding " + } + ] + }, + { + "id": "MISS_1_1_801_0", + "timepoint": 801, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 801, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1815_0", + "timepoint": 1815, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1815, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,needle,holding; assistant,operating table,closeto; needle,patient,touching; assistant,syringe,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1112_0", + "timepoint": 1112, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1112, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; forceps,cotton,touching; assistant,operating table,closeto; circulator,ultrasound machine,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,antiseptic,holding; circulator,ultrasound machine,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_1039_0", + "timepoint": 1039, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1039, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2413_0", + "timepoint": 2413, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2413, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,holding; syringe,needle,touching; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_325_0", + "timepoint": 325, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 325, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_992_0", + "timepoint": 992, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 992, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_237_0", + "timepoint": 237, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 237, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,instrument table,looking; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_972_0", + "timepoint": 972, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 972, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching; cotton,patient,touching; assistant,forceps,holding; assistant,patient,disinfection; assistant,antiseptic,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_349_0", + "timepoint": 349, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 349, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; anesthetist,health monitor,looking; assistant,operating table,closeto; anesthetist,operating table,closeto; cotton,patient,touching; head surgeon,scalpel,holding; assistant,forceps,holding; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,patient,aspirating " + } + ] + }, + { + "id": "Ultrasound_5_6_2663_0", + "timepoint": 2663, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2663, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2674_0", + "timepoint": 2674, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2674, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; needle,patient,touching; assistant,syringe,injecting; syringe,needle,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_382_0", + "timepoint": 382, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 382, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_933_0", + "timepoint": 933, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 933, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2733_0", + "timepoint": 2733, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2733, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,bin,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2975_0", + "timepoint": 2975, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2975, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; tissue paper,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2828_0", + "timepoint": 2828, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2828, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; head surgeon,ultrasound machine,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_198_0", + "timepoint": 198, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 198, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,looking; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_687_0", + "timepoint": 687, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 687, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_413_0", + "timepoint": 413, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 413, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_1372_0", + "timepoint": 1372, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1372, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,body marker,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_970_0", + "timepoint": 970, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 970, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,microscope,manipulating; anesthetist,operating table,closeto; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_133_0", + "timepoint": 133, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 133, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_846_0", + "timepoint": 846, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 846, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_818_0", + "timepoint": 818, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 818, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_3_6_22_0", + "timepoint": 22, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 22, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_3_6_174_0", + "timepoint": 174, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 174, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,forceps,holding; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; curette,patient,touching; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "MISS_1_1_581_0", + "timepoint": 581, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 581, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering; circulator,operating table,manipulating " + } + ] + }, + { + "id": "MISS_1_1_1106_0", + "timepoint": 1106, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1106, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_107_0", + "timepoint": 107, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 107, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1372_0", + "timepoint": 1372, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1372, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1231_0", + "timepoint": 1231, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1231, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; patient,operating table,lyingon; assistant,antiseptic,holding; assistant,forceps,holding; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2466_0", + "timepoint": 2466, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2466, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,ultrasound machine,looking; head surgeon,operating table,closeto; assistant,needle,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1534_0", + "timepoint": 1534, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1534, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1631_0", + "timepoint": 1631, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1631, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,instrument table,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_2311_0", + "timepoint": 2311, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2311, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_895_0", + "timepoint": 895, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 895, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,microscope,manipulating; circulator,microscope,manipulating " + } + ] + }, + { + "id": "MISS_1_1_216_0", + "timepoint": 216, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 216, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_62_0", + "timepoint": 62, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 62, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3317_0", + "timepoint": 3317, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3317, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,bin,looking; head surgeon,operating table,closeto; assistant,bin,closeto; assistant,tissue paper,dropping " + } + ] + }, + { + "id": "Ultrasound_3_4_2943_0", + "timepoint": 2943, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2943, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; head surgeon,operating table,closeto; patient,operating table,lyingon; tissue paper,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1920_0", + "timepoint": 1920, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1920, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1496_0", + "timepoint": 1496, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1496, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_5_6_971_0", + "timepoint": 971, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 971, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1631_0", + "timepoint": 1631, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1631, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,inserting; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_962_0", + "timepoint": 962, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 962, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; cotton,patient,touching; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,antiseptic,holding; assistant,patient,disinfection; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2495_0", + "timepoint": 2495, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2495, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2013_0", + "timepoint": 2013, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2013, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; syringe,needle,touching; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,needle,holding; assistant,syringe,injecting; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_695_0", + "timepoint": 695, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 695, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,operating table,closeto; assistant,curette,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,instrument table,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2759_0", + "timepoint": 2759, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2759, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; assistant,needle,holding; assistant,syringe,injecting; syringe,needle,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2326_0", + "timepoint": 2326, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2326, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_28_0", + "timepoint": 28, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 28, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2042_0", + "timepoint": 2042, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2042, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_2102_0", + "timepoint": 2102, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2102, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,forceps,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,disinfection; head surgeon,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2227_0", + "timepoint": 2227, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2227, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,syringe,injecting; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_344_0", + "timepoint": 344, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 344, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1535_0", + "timepoint": 1535, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1535, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,needle,inserting; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1539_0", + "timepoint": 1539, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1539, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1653_0", + "timepoint": 1653, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1653, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2651_0", + "timepoint": 2651, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2651, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_3_4_1400_0", + "timepoint": 1400, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1400, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,body marker,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_601_0", + "timepoint": 601, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 601, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_95_0", + "timepoint": 95, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 95, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; curette,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2442_0", + "timepoint": 2442, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2442, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; needle,patient,touching; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2491_0", + "timepoint": 2491, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2491, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,needle,holding; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,needle,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_339_0", + "timepoint": 339, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 339, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,looking; patient,operating table,closeto; circulator,patient,closeto " + } + ] + }, + { + "id": "MISS_3_6_239_0", + "timepoint": 239, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 239, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,forceps,holding; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; anesthetist,health monitor,looking; patient,operating table,lyingon; curette,patient,touching; forceps,patient,touching; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,operating table,closeto; head surgeon,curette,holding " + } + ] + }, + { + "id": "MISS_1_1_1657_0", + "timepoint": 1657, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1657, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_474_0", + "timepoint": 474, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 474, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1046_0", + "timepoint": 1046, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1046, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; assistant,antiseptic,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_280_0", + "timepoint": 280, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 280, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_954_0", + "timepoint": 954, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 954, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,patient,disinfection; assistant,operating table,closeto; cotton,patient,touching; assistant,antiseptic,holding; head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2292_0", + "timepoint": 2292, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2292, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; syringe,needle,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_2_2_775_0", + "timepoint": 775, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 775, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "MISS_1_1_1362_0", + "timepoint": 1362, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1362, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,operating table,looking; assistant,instrument table,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,patient,anaesthetising " + } + ] + }, + { + "id": "MISS_3_6_1020_0", + "timepoint": 1020, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1020, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,patient,aspirating; cotton,patient,touching; anesthetist,health monitor,looking; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1298_0", + "timepoint": 1298, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1298, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " body marker,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling; assistant,body marker,applying; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1753_0", + "timepoint": 1753, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1753, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; assistant,instrument table,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_937_0", + "timepoint": 937, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 937, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_357_0", + "timepoint": 357, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 357, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,looking; patient,operating room,entering; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2715_0", + "timepoint": 2715, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2715, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,bin,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1539_0", + "timepoint": 1539, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1539, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,manipulating; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,patient,anaesthetising; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1307_0", + "timepoint": 1307, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1307, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_379_0", + "timepoint": 379, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 379, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; curette,patient,touching; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; forceps,patient,touching; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,forceps,holding; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2396_0", + "timepoint": 2396, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2396, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,syringe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2921_0", + "timepoint": 2921, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2921, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_116_0", + "timepoint": 116, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 116, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2258_0", + "timepoint": 2258, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2258, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; syringe,needle,touching; assistant,syringe,removing; head surgeon,vertebrae,scanning; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3214_0", + "timepoint": 3214, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3214, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,tissue paper,applying; assistant,operating table,closeto; tissue paper,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1851_0", + "timepoint": 1851, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1851, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,antiseptic,applying; assistant,patient,disinfection; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1864_0", + "timepoint": 1864, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1864, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,syringe,injecting; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2869_0", + "timepoint": 2869, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2869, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; syringe,needle,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; assistant,syringe,removing; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1707_0", + "timepoint": 1707, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1707, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,operating table,closeto; assistant,tissue mark,looking; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_895_0", + "timepoint": 895, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 895, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1126_0", + "timepoint": 1126, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1126, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,closeto; assistant,forceps,holding; circulator,ultrasound machine,manipulating; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,antiseptic,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_393_0", + "timepoint": 393, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 393, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1625_0", + "timepoint": 1625, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1625, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2192_0", + "timepoint": 2192, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2192, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1478_0", + "timepoint": 1478, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1478, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,ultrasound probe,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2129_0", + "timepoint": 2129, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2129, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; assistant,syringe,injecting; assistant,needle,holding; syringe,needle,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1649_0", + "timepoint": 1649, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1649, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2546_0", + "timepoint": 2546, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2546, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2103_0", + "timepoint": 2103, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2103, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; syringe,needle,touching; patient,operating table,lyingon; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_3_6_377_0", + "timepoint": 377, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 377, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; curette,patient,touching; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; forceps,patient,touching; assistant,operating table,closeto; head surgeon,curette,holding; head surgeon,microscope eye,looking; head surgeon,forceps,holding; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_89_0", + "timepoint": 89, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 89, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_2746_0", + "timepoint": 2746, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2746, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1913_0", + "timepoint": 1913, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1913, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; patient,operating table,lyingon; assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2724_0", + "timepoint": 2724, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2724, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,bin,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_318_0", + "timepoint": 318, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 318, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1692_0", + "timepoint": 1692, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1692, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_4_5_1233_0", + "timepoint": 1233, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1233, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,ultrasound probe,looking; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1572_0", + "timepoint": 1572, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1572, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1044_0", + "timepoint": 1044, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1044, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2634_0", + "timepoint": 2634, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2634, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_3_4_1932_0", + "timepoint": 1932, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1932, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; needle,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; assistant,instrument table,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2671_0", + "timepoint": 2671, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2671, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,tissue paper,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_44_0", + "timepoint": 44, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 44, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2874_0", + "timepoint": 2874, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2874, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; syringe,needle,touching; assistant,syringe,removing; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2055_0", + "timepoint": 2055, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2055, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1027_0", + "timepoint": 1027, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1027, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; cotton,patient,touching; assistant,antiseptic,holding; patient,operating table,lyingon; assistant,patient,disinfection; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1117_0", + "timepoint": 1117, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1117, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,needle,inserting; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_638_0", + "timepoint": 638, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 638, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_241_0", + "timepoint": 241, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 241, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; forceps,patient,touching; assistant,operating table,closeto; head surgeon,curette,holding; anesthetist,health monitor,looking; curette,patient,touching; patient,operating table,lyingon; head surgeon,herbal disk,removing; head surgeon,forceps,holding; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1545_0", + "timepoint": 1545, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1545, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,needle,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1514_0", + "timepoint": 1514, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1514, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,patient,anaesthetising; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope controller,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_2748_0", + "timepoint": 2748, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2748, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,instrument table,closeto; head surgeon,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2939_0", + "timepoint": 2939, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2939, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; tissue paper,patient,touching; head surgeon,tissue paper,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2378_0", + "timepoint": 2378, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2378, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,syringe,injecting; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1456_0", + "timepoint": 1456, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1456, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1880_0", + "timepoint": 1880, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1880, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,applying; assistant,patient,disinfection; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1735_0", + "timepoint": 1735, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1735, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,holding; assistant,instrument table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_2219_0", + "timepoint": 2219, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2219, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,patient,disinfection; forceps,cotton,touching; assistant,forceps,holding; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_662_0", + "timepoint": 662, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 662, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_879_0", + "timepoint": 879, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 879, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "MISS_1_1_1178_0", + "timepoint": 1178, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1178, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating room,entering; patient,operating table,lyingon; assistant,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_2_2_986_0", + "timepoint": 986, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 986, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; patient,operating table,lyingon; cotton,patient,touching; forceps,cotton,touching; assistant,antiseptic,holding; assistant,operating table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2841_0", + "timepoint": 2841, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2841, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_719_0", + "timepoint": 719, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 719, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_701_0", + "timepoint": 701, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 701, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_526_0", + "timepoint": 526, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 526, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1601_0", + "timepoint": 1601, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1601, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_125_0", + "timepoint": 125, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 125, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,ultrasound gel,applying; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_3_6_401_0", + "timepoint": 401, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 401, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; head surgeon,curette,holding; head surgeon,forceps,holding; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; head surgeon,microscope eye,looking; curette,patient,touching; assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2901_0", + "timepoint": 2901, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2901, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,needle,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,syringe,removing; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2581_0", + "timepoint": 2581, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2581, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,syringe,injecting; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3187_0", + "timepoint": 3187, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3187, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; tissue paper,patient,touching; patient,operating table,lyingon; assistant,tissue paper,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_636_0", + "timepoint": 636, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 636, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,curette,holding; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_834_0", + "timepoint": 834, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 834, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2008_0", + "timepoint": 2008, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2008, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; syringe,needle,touching; ultrasound probe,patient,touching; assistant,syringe,injecting; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2328_0", + "timepoint": 2328, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2328, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; syringe,needle,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_404_0", + "timepoint": 404, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 404, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,instrument table,looking; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1186_0", + "timepoint": 1186, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1186, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,inserting; assistant,ultrasound probe,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1329_0", + "timepoint": 1329, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1329, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,body marker,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_203_0", + "timepoint": 203, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 203, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_915_0", + "timepoint": 915, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 915, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; cotton,patient,touching; assistant,patient,disinfection; patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; assistant,antiseptic,holding " + } + ] + }, + { + "id": "MISS_2_2_619_0", + "timepoint": 619, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 619, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; head surgeon,microscope screen,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,microscope controller,controlling; head surgeon,operating table,closeto; assistant,microscope screen,looking; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1607_0", + "timepoint": 1607, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1607, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_417_0", + "timepoint": 417, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 417, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,patient,closeto; circulator,operating table,closeto; assistant,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_2_2_54_0", + "timepoint": 54, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 54, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_3_6_1062_0", + "timepoint": 1062, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1062, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; anesthetist,operating table,closeto; assistant,forceps,holding; anesthetist,health monitor,looking; cotton,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,microscope eye,looking; assistant,patient,aspirating; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_601_0", + "timepoint": 601, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 601, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_282_0", + "timepoint": 282, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 282, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; curette,patient,touching; head surgeon,forceps,holding; head surgeon,herbal disk,removing; patient,operating table,lyingon; head surgeon,microscope eye,looking; forceps,patient,touching; anesthetist,health monitor,looking; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_1528_0", + "timepoint": 1528, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1528, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,patient,anaesthetising; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2159_0", + "timepoint": 2159, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2159, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; anesthetist,operating table,closeto; cotton,patient,touching; assistant,patient,disinfection; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1185_0", + "timepoint": 1185, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1185, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2091_0", + "timepoint": 2091, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2091, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,vertebrae,scanning; syringe,needle,touching; needle,patient,touching; head surgeon,needle,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_472_0", + "timepoint": 472, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 472, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering; anesthetist,operating table,manipulating; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_1899_0", + "timepoint": 1899, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1899, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,needle,holding; needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2505_0", + "timepoint": 2505, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2505, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; needle,patient,touching; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_391_0", + "timepoint": 391, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 391, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_3079_0", + "timepoint": 3079, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3079, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_620_0", + "timepoint": 620, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 620, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1398_0", + "timepoint": 1398, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1398, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; assistant,needle,removing; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_880_0", + "timepoint": 880, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 880, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,antiseptic,applying; assistant,forceps,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_5_6_2799_0", + "timepoint": 2799, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2799, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; syringe,needle,touching; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_1152_0", + "timepoint": 1152, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1152, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; patient,operating table,lyingon; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_3_6_1076_0", + "timepoint": 1076, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1076, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; cotton,patient,touching; anesthetist,operating table,closeto; anesthetist,health monitor,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,aspirating; forceps,cotton,touching; head surgeon,microscope eye,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2031_0", + "timepoint": 2031, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2031, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,inserting; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,syringe,holding; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_615_0", + "timepoint": 615, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 615, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_149_0", + "timepoint": 149, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 149, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,patient,manipulating; head surgeon,ultrasound machine,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_76_0", + "timepoint": 76, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 76, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_399_0", + "timepoint": 399, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 399, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_31_0", + "timepoint": 31, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 31, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating room,entering " + } + ] + }, + { + "id": "MISS_1_1_1453_0", + "timepoint": 1453, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1453, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,manipulating; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3162_0", + "timepoint": 3162, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3162, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; tissue paper,patient,touching; assistant,tissue paper,applying " + } + ] + }, + { + "id": "MISS_1_1_2111_0", + "timepoint": 2111, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2111, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,patient,disinfection; cotton,patient,touching; assistant,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1114_0", + "timepoint": 1114, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1114, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1224_0", + "timepoint": 1224, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1224, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,antiseptic,holding; patient,operating table,lyingon; assistant,operating table,closeto; circulator,ultrasound machine,closeto; forceps,cotton,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2506_0", + "timepoint": 2506, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2506, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1966_0", + "timepoint": 1966, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1966, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,needle,holding; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_948_0", + "timepoint": 948, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 948, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,assistant,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1206_0", + "timepoint": 1206, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1206, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,body marker,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_14_0", + "timepoint": 14, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 14, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_631_0", + "timepoint": 631, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 631, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2080_0", + "timepoint": 2080, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2080, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,syringe,injecting; needle,patient,touching; ultrasound probe,patient,touching; assistant,operating table,closeto; syringe,needle,touching; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_482_0", + "timepoint": 482, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 482, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; head surgeon,curette,holding; head surgeon,herbal disk,removing; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; scalpel,patient,touching; head surgeon,microscope eye,looking; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3319_0", + "timepoint": 3319, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3319, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,bin,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1122_0", + "timepoint": 1122, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1122, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1758_0", + "timepoint": 1758, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1758, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; assistant,syringe,holding; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3056_0", + "timepoint": 3056, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3056, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,tissue paper,holding " + } + ] + }, + { + "id": "MISS_3_6_686_0", + "timepoint": 686, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 686, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,instrument table,looking; assistant,curette,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2310_0", + "timepoint": 2310, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2310, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; syringe,needle,touching; needle,patient,touching; head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_2_2_651_0", + "timepoint": 651, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 651, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,microscope screen,looking; head surgeon,operating table,closeto; head surgeon,microscope screen,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,controlling " + } + ] + }, + { + "id": "MISS_1_1_1058_0", + "timepoint": 1058, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1058, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_723_0", + "timepoint": 723, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 723, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,microscope screen,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2124_0", + "timepoint": 2124, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2124, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,needle,holding; syringe,needle,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,injecting; needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2721_0", + "timepoint": 2721, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2721, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,syringe,injecting; assistant,needle,holding; syringe,needle,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_3172_0", + "timepoint": 3172, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3172, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_3315_0", + "timepoint": 3315, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3315, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,bin,closeto; assistant,bin,looking; patient,operating table,lyingon; assistant,tissue paper,dropping " + } + ] + }, + { + "id": "Ultrasound_4_5_1955_0", + "timepoint": 1955, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1955, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,syringe,injecting; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1253_0", + "timepoint": 1253, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1253, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; forceps,cotton,touching; assistant,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1833_0", + "timepoint": 1833, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1833, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_927_0", + "timepoint": 927, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 927, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_238_0", + "timepoint": 238, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 238, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_365_0", + "timepoint": 365, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 365, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; anesthetist,operating table,manipulating; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "MISS_3_6_711_0", + "timepoint": 711, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 711, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; assistant,curette,holding; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1160_0", + "timepoint": 1160, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1160, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2288_0", + "timepoint": 2288, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2288, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; needle,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2404_0", + "timepoint": 2404, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2404, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; assistant,instrument table,looking; assistant,syringe,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_111_0", + "timepoint": 111, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 111, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; patient,operating table,lyingon; assistant,instrument table,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2543_0", + "timepoint": 2543, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2543, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; assistant,syringe,injecting; assistant,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_950_0", + "timepoint": 950, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 950, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_3014_0", + "timepoint": 3014, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3014, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,tissue paper,holding; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_968_0", + "timepoint": 968, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 968, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1757_0", + "timepoint": 1757, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1757, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,syringe,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,instrument table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1258_0", + "timepoint": 1258, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1258, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2728_0", + "timepoint": 2728, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2728, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,bin,closeto " + } + ] + }, + { + "id": "MISS_1_1_447_0", + "timepoint": 447, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 447, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; anesthetist,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_694_0", + "timepoint": 694, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 694, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; circulator,ultrasound machine,looking; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_1038_0", + "timepoint": 1038, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1038, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; patient,operating table,lyingon; cotton,patient,touching; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,antiseptic,holding; assistant,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_3_4_2541_0", + "timepoint": 2541, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2541, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; syringe,needle,touching; patient,operating table,lyingon; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,needle,holding; assistant,syringe,injecting; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1250_0", + "timepoint": 1250, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1250, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,patient,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2301_0", + "timepoint": 2301, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2301, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,holding; needle,patient,touching; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_1_1_388_0", + "timepoint": 388, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 388, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,operating table,manipulating; patient,operating table,lyingon; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1495_0", + "timepoint": 1495, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1495, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,ultrasound probe,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_4_5_2472_0", + "timepoint": 2472, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2472, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,operating table,closeto; assistant,instrument table,looking; head surgeon,ultrasound machine,looking; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1248_0", + "timepoint": 1248, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1248, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; head surgeon,operating table,closeto; assistant,antiseptic,holding; forceps,cotton,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_76_0", + "timepoint": 76, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 76, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2215_0", + "timepoint": 2215, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2215, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,needle,holding; syringe,needle,touching; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2360_0", + "timepoint": 2360, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2360, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1100_0", + "timepoint": 1100, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1100, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,antiseptic,holding; circulator,ultrasound machine,closeto; patient,operating table,lyingon; forceps,cotton,touching; head surgeon,operating table,closeto; circulator,ultrasound machine,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_978_0", + "timepoint": 978, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 978, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,antiseptic,holding; assistant,patient,disinfection; forceps,cotton,touching; assistant,forceps,holding; assistant,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2271_0", + "timepoint": 2271, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2271, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,needle,holding; assistant,operating table,closeto; needle,patient,touching; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1924_0", + "timepoint": 1924, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1924, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,antiseptic,applying; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_4_5_2678_0", + "timepoint": 2678, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2678, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_727_0", + "timepoint": 727, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 727, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_5_0", + "timepoint": 5, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 5, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering " + } + ] + }, + { + "id": "MISS_3_6_312_0", + "timepoint": 312, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 312, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,curette,holding; forceps,patient,touching; anesthetist,health monitor,looking; head surgeon,forceps,holding; anesthetist,anesthesia equipment,controlling; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_302_0", + "timepoint": 302, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 302, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,patient,closeto " + } + ] + }, + { + "id": "MISS_1_1_320_0", + "timepoint": 320, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 320, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; anesthetist,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1614_0", + "timepoint": 1614, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1614, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,ultrasound screen,looking; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_3_4_2280_0", + "timepoint": 2280, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2280, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; needle,patient,touching; assistant,syringe,injecting; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,ultrasound screen,looking; assistant,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_673_0", + "timepoint": 673, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 673, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_761_0", + "timepoint": 761, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 761, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2761_0", + "timepoint": 2761, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2761, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,syringe,injecting; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_878_0", + "timepoint": 878, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 878, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_2_2_467_0", + "timepoint": 467, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 467, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,scalpel,holding; patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_714_0", + "timepoint": 714, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 714, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1006_0", + "timepoint": 1006, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1006, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1140_0", + "timepoint": 1140, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1140, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,health monitor,looking; assistant,instrument table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2737_0", + "timepoint": 2737, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2737, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,needle,holding; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; assistant,instrument table,looking; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; assistant,syringe,holding " + } + ] + }, + { + "id": "MISS_1_1_2138_0", + "timepoint": 2138, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2138, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; cotton,patient,touching; assistant,operating table,closeto; assistant,forceps,holding; assistant,patient,disinfection; patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_2_2_686_0", + "timepoint": 686, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 686, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope screen,looking; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope controller,controlling; assistant,microscope screen,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2251_0", + "timepoint": 2251, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2251, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,forceps,holding; cotton,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2503_0", + "timepoint": 2503, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2503, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; needle,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_40_0", + "timepoint": 40, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 40, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1694_0", + "timepoint": 1694, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1694, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,needle,inserting; assistant,tissue mark,looking; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_526_0", + "timepoint": 526, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 526, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1869_0", + "timepoint": 1869, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1869, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,syringe,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_513_0", + "timepoint": 513, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 513, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_501_0", + "timepoint": 501, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 501, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_610_0", + "timepoint": 610, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 610, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_839_0", + "timepoint": 839, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 839, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1326_0", + "timepoint": 1326, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1326, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; patient,operating table,lyingon; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1705_0", + "timepoint": 1705, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1705, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; needle,patient,touching; assistant,tissue mark,looking; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,needle,inserting " + } + ] + }, + { + "id": "MISS_2_2_665_0", + "timepoint": 665, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 665, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,microscope controller,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,microscope screen,looking; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1624_0", + "timepoint": 1624, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1624, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,tissue mark,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_663_0", + "timepoint": 663, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 663, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; assistant,curette,holding; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,scalpel,holding; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_620_0", + "timepoint": 620, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 620, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_2_2_88_0", + "timepoint": 88, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 88, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_228_0", + "timepoint": 228, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 228, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1373_0", + "timepoint": 1373, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1373, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2261_0", + "timepoint": 2261, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2261, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; syringe,needle,touching; assistant,needle,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2693_0", + "timepoint": 2693, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2693, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,syringe,injecting; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1123_0", + "timepoint": 1123, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1123, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,microscope eye,looking; head surgeon,microscope eye,looking; assistant,forceps,holding; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_408_0", + "timepoint": 408, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 408, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,touching; circulator,operating table,closeto; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1741_0", + "timepoint": 1741, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1741, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; needle,patient,touching; assistant,instrument table,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1278_0", + "timepoint": 1278, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1278, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,inserting; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_48_0", + "timepoint": 48, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 48, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "MISS_2_2_452_0", + "timepoint": 452, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 452, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; assistant,forceps,holding; head surgeon,scalpel,holding; assistant,instrument table,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,cotton,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_210_0", + "timepoint": 210, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 210, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,ultrasound gel,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_661_0", + "timepoint": 661, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 661, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; patient,operating table,lyingon; assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2441_0", + "timepoint": 2441, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2441, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound machine,looking; needle,patient,touching; assistant,needle,removing; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_264_0", + "timepoint": 264, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 264, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,patient,cutting; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,forceps,holding; anesthetist,operating table,closeto; scalpel,patient,touching; patient,operating table,lyingon; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "MISS_3_6_952_0", + "timepoint": 952, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 952, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; anesthetist,operating table,closeto; assistant,patient,aspirating; head surgeon,microscope eye,looking; cotton,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2727_0", + "timepoint": 2727, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2727, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,needle,holding; needle,patient,touching; assistant,operating table,closeto; assistant,instrument table,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,syringe,holding " + } + ] + }, + { + "id": "MISS_1_1_140_0", + "timepoint": 140, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 140, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_632_0", + "timepoint": 632, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 632, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_468_0", + "timepoint": 468, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 468, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; anesthetist,health monitor,looking; patient,operating table,lyingon; assistant,scalpel,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1160_0", + "timepoint": 1160, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1160, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_704_0", + "timepoint": 704, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 704, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_643_0", + "timepoint": 643, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 643, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1923_0", + "timepoint": 1923, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1923, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,patient,disinfection; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,antiseptic,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_1869_0", + "timepoint": 1869, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1869, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,syringe,injecting; needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2006_0", + "timepoint": 2006, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2006, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,ultrasound probe,looking; patient,operating table,lyingon; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_3_4_2261_0", + "timepoint": 2261, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2261, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,syringe,injecting; assistant,needle,holding; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1222_0", + "timepoint": 1222, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1222, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; circulator,ultrasound machine,closeto; assistant,antiseptic,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching; circulator,ultrasound machine,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_475_0", + "timepoint": 475, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 475, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2580_0", + "timepoint": 2580, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2580, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_77_0", + "timepoint": 77, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 77, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,antiseptic,looking; circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_3_6_75_0", + "timepoint": 75, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 75, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; head surgeon,curette,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1628_0", + "timepoint": 1628, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1628, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_22_0", + "timepoint": 22, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 22, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_404_0", + "timepoint": 404, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 404, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_853_0", + "timepoint": 853, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 853, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,body marker,applying; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; body marker,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_215_0", + "timepoint": 215, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 215, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; head surgeon,operating table,closeto; head surgeon,forceps,holding; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,herbal disk,removing; patient,operating table,lyingon; curette,patient,touching; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_3230_0", + "timepoint": 3230, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3230, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; tissue paper,patient,touching; patient,operating table,lyingon; assistant,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_2965_0", + "timepoint": 2965, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2965, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,patient,scanning; head surgeon,operating table,closeto; assistant,needle,removing; assistant,syringe,holding " + } + ] + }, + { + "id": "MISS_1_1_1594_0", + "timepoint": 1594, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1594, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1878_0", + "timepoint": 1878, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1878, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,antiseptic,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "MISS_3_6_251_0", + "timepoint": 251, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 251, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,health monitor,looking; head surgeon,forceps,holding; forceps,patient,touching; anesthetist,operating table,closeto; curette,patient,touching; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,curette,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1441_0", + "timepoint": 1441, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1441, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,patient,anaesthetising; head surgeon,microscope controller,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1940_0", + "timepoint": 1940, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1940, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,syringe,injecting; assistant,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1774_0", + "timepoint": 1774, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1774, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,syringe,holding; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2678_0", + "timepoint": 2678, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2678, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_993_0", + "timepoint": 993, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 993, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1269_0", + "timepoint": 1269, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1269, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_1345_0", + "timepoint": 1345, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1345, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; anesthetist,patient,anaesthetising; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_940_0", + "timepoint": 940, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 940, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; cotton,patient,touching; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,patient,aspirating; assistant,forceps,holding; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2705_0", + "timepoint": 2705, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2705, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,syringe,removing; assistant,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1473_0", + "timepoint": 1473, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1473, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1776_0", + "timepoint": 1776, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1776, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue mark,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; assistant,needle,inserting; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1873_0", + "timepoint": 1873, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1873, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,applying; assistant,patient,disinfection; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1806_0", + "timepoint": 1806, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1806, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; assistant,syringe,holding; head surgeon,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1133_0", + "timepoint": 1133, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1133, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1722_0", + "timepoint": 1722, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1722, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1204_0", + "timepoint": 1204, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1204, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,ultrasound probe,looking; assistant,needle,inserting; head surgeon,operating table,closeto; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_411_0", + "timepoint": 411, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 411, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_361_0", + "timepoint": 361, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 361, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; curette,patient,touching; forceps,patient,touching; head surgeon,microscope eye,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,curette,holding " + } + ] + }, + { + "id": "MISS_1_1_246_0", + "timepoint": 246, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 246, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_754_0", + "timepoint": 754, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 754, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1727_0", + "timepoint": 1727, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1727, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1007_0", + "timepoint": 1007, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1007, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,disinfection; cotton,patient,touching; assistant,antiseptic,holding; patient,operating table,lyingon; forceps,cotton,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_2_2_460_0", + "timepoint": 460, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 460, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; patient,operating table,lyingon; anesthetist,health monitor,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2452_0", + "timepoint": 2452, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2452, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_409_0", + "timepoint": 409, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 409, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1808_0", + "timepoint": 1808, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1808, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1294_0", + "timepoint": 1294, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1294, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2450_0", + "timepoint": 2450, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2450, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,syringe,holding; patient,operating table,lyingon; head surgeon,needle,holding; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3068_0", + "timepoint": 3068, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3068, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,tissue paper,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1828_0", + "timepoint": 1828, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1828, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,disinfection; patient,operating table,lyingon; assistant,antiseptic,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_713_0", + "timepoint": 713, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 713, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,closeto; head surgeon,microscope eye,looking; assistant,instrument table,looking; assistant,curette,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_320_0", + "timepoint": 320, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 320, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,looking; patient,operating room,entering; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2169_0", + "timepoint": 2169, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2169, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon; assistant,syringe,injecting; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_895_0", + "timepoint": 895, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 895, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2559_0", + "timepoint": 2559, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2559, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_3_4_1029_0", + "timepoint": 1029, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1029, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,body marker,holding; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2695_0", + "timepoint": 2695, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2695, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,syringe,injecting; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1759_0", + "timepoint": 1759, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1759, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,instrument table,closeto; assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3021_0", + "timepoint": 3021, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3021, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2981_0", + "timepoint": 2981, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2981, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,patient,scanning; assistant,needle,holding; assistant,syringe,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1113_0", + "timepoint": 1113, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1113, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; needle,patient,touching; assistant,needle,inserting; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_268_0", + "timepoint": 268, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 268, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_414_0", + "timepoint": 414, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 414, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_3_6_26_0", + "timepoint": 26, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 26, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_496_0", + "timepoint": 496, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 496, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " scalpel,patient,touching; anesthetist,operating table,closeto; curette,patient,touching; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,scalpel,holding; head surgeon,herbal disk,removing; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_346_0", + "timepoint": 346, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 346, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2048_0", + "timepoint": 2048, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2048, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; ultrasound probe,patient,touching; syringe,needle,touching; patient,operating table,lyingon; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_683_0", + "timepoint": 683, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 683, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,curette,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2964_0", + "timepoint": 2964, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2964, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,patient,scanning; assistant,syringe,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_177_0", + "timepoint": 177, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 177, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; forceps,cotton,touching; anesthetist,operating table,closeto; head surgeon,scalpel,holding; head surgeon,operating table,closeto; scalpel,patient,touching; head surgeon,patient,cutting " + } + ] + }, + { + "id": "Ultrasound_5_6_2796_0", + "timepoint": 2796, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2796, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; syringe,needle,touching; patient,operating table,lyingon; assistant,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_3091_0", + "timepoint": 3091, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3091, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,tissue paper,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1110_0", + "timepoint": 1110, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1110, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; patient,operating table,lyingon; assistant,operating table,closeto; circulator,ultrasound machine,closeto; assistant,antiseptic,holding; head surgeon,operating table,closeto; forceps,cotton,touching; circulator,ultrasound machine,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_742_0", + "timepoint": 742, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 742, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_139_0", + "timepoint": 139, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 139, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1462_0", + "timepoint": 1462, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1462, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2129_0", + "timepoint": 2129, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2129, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; head surgeon,operating table,closeto; anesthetist,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; assistant,forceps,holding; assistant,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_2_2_570_0", + "timepoint": 570, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 570, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2196_0", + "timepoint": 2196, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2196, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,needle,holding; assistant,ultrasound screen,looking; needle,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_188_0", + "timepoint": 188, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 188, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2376_0", + "timepoint": 2376, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2376, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,syringe,holding; needle,patient,touching; patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_2296_0", + "timepoint": 2296, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2296, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; forceps,cotton,touching; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2907_0", + "timepoint": 2907, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2907, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_966_0", + "timepoint": 966, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 966, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_130_0", + "timepoint": 130, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 130, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_20_0", + "timepoint": 20, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 20, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering; circulator,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_36_0", + "timepoint": 36, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 36, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_2_2_379_0", + "timepoint": 379, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 379, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; cotton,patient,touching; assistant,operating table,closeto; assistant,patient,aspirating; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "MISS_1_1_2033_0", + "timepoint": 2033, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2033, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,patient,disinfection; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2880_0", + "timepoint": 2880, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2880, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,syringe,removing; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,needle,holding; assistant,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; syringe,needle,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_749_0", + "timepoint": 749, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 749, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_2_2_463_0", + "timepoint": 463, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 463, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1270_0", + "timepoint": 1270, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1270, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " body marker,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,body marker,applying; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1867_0", + "timepoint": 1867, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1867, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; syringe,needle,touching; assistant,syringe,injecting; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_658_0", + "timepoint": 658, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 658, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_280_0", + "timepoint": 280, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 280, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1708_0", + "timepoint": 1708, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1708, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; head surgeon,operating table,closeto; assistant,tissue mark,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2064_0", + "timepoint": 2064, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2064, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,closeto; forceps,cotton,touching; assistant,patient,disinfection; head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1620_0", + "timepoint": 1620, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1620, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_4_5_269_0", + "timepoint": 269, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 269, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_1_1_1623_0", + "timepoint": 1623, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1623, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2774_0", + "timepoint": 2774, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2774, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1394_0", + "timepoint": 1394, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1394, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_481_0", + "timepoint": 481, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 481, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1653_0", + "timepoint": 1653, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1653, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_111_0", + "timepoint": 111, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 111, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2476_0", + "timepoint": 2476, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2476, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,syringe,holding " + } + ] + }, + { + "id": "MISS_3_6_155_0", + "timepoint": 155, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 155, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; curette,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,curette,holding; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1355_0", + "timepoint": 1355, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1355, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2025_0", + "timepoint": 2025, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2025, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,inserting; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2576_0", + "timepoint": 2576, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2576, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,patient,manipulating; assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3024_0", + "timepoint": 3024, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3024, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,applying; tissue paper,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_58_0", + "timepoint": 58, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 58, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,scalpel,looking; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2646_0", + "timepoint": 2646, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2646, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,tissue paper,applying; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_724_0", + "timepoint": 724, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 724, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_206_0", + "timepoint": 206, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 206, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_338_0", + "timepoint": 338, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 338, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,ultrasound gel,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_12_0", + "timepoint": 12, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 12, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3079_0", + "timepoint": 3079, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3079, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,tissue paper,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2462_0", + "timepoint": 2462, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2462, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; syringe,needle,touching; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; assistant,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_2085_0", + "timepoint": 2085, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2085, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,disinfection; assistant,forceps,holding; patient,operating table,lyingon; forceps,cotton,touching; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2316_0", + "timepoint": 2316, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2316, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; syringe,needle,touching; assistant,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_810_0", + "timepoint": 810, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 810, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,body marker,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1019_0", + "timepoint": 1019, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1019, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; cotton,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching; assistant,forceps,holding; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_111_0", + "timepoint": 111, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 111, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "MISS_3_6_563_0", + "timepoint": 563, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 563, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; patient,operating table,lyingon; scalpel,patient,touching; head surgeon,scalpel,holding; curette,patient,touching; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_278_0", + "timepoint": 278, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 278, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; head surgeon,scalpel,holding; forceps,cotton,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_574_0", + "timepoint": 574, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 574, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2283_0", + "timepoint": 2283, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2283, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; assistant,syringe,holding; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_761_0", + "timepoint": 761, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 761, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; circulator,microscope,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1305_0", + "timepoint": 1305, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1305, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_619_0", + "timepoint": 619, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 619, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,health monitor,looking; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; curette,patient,touching; scalpel,patient,touching; head surgeon,scalpel,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1964_0", + "timepoint": 1964, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1964, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning; head surgeon,needle,holding; needle,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1378_0", + "timepoint": 1378, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1378, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,operating table,closeto; assistant,needle,removing; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_181_0", + "timepoint": 181, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 181, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1370_0", + "timepoint": 1370, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1370, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2988_0", + "timepoint": 2988, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2988, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,patient,scanning; assistant,instrument table,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3273_0", + "timepoint": 3273, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3273, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; tissue paper,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1391_0", + "timepoint": 1391, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1391, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_814_0", + "timepoint": 814, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 814, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2400_0", + "timepoint": 2400, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2400, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,needle,removing; needle,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1098_0", + "timepoint": 1098, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1098, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,body marker,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_229_0", + "timepoint": 229, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 229, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_2_2_721_0", + "timepoint": 721, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 721, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,microscope screen,looking; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_2028_0", + "timepoint": 2028, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2028, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,forceps,holding; assistant,patient,disinfection; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2303_0", + "timepoint": 2303, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2303, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2295_0", + "timepoint": 2295, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2295, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; forceps,cotton,touching; assistant,forceps,holding; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2394_0", + "timepoint": 2394, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2394, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,syringe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_736_0", + "timepoint": 736, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 736, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "MISS_3_6_525_0", + "timepoint": 525, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 525, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; scalpel,patient,touching; curette,patient,touching; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; patient,operating table,lyingon; head surgeon,herbal disk,removing; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2458_0", + "timepoint": 2458, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2458, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound machine,looking; assistant,instrument table,looking; assistant,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1031_0", + "timepoint": 1031, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1031, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2245_0", + "timepoint": 2245, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2245, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; needle,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_3_6_78_0", + "timepoint": 78, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 78, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,curette,holding; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_2_2_540_0", + "timepoint": 540, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 540, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope controller,controlling; assistant,operating table,closeto; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_173_0", + "timepoint": 173, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 173, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1203_0", + "timepoint": 1203, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1203, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,body marker,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_902_0", + "timepoint": 902, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 902, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,microscope,manipulating; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2240_0", + "timepoint": 2240, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2240, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; syringe,needle,touching; needle,patient,touching; ultrasound probe,patient,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_5_6_3244_0", + "timepoint": 3244, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3244, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,tissue paper,applying; tissue paper,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1257_0", + "timepoint": 1257, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1257, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,body marker,applying; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2443_0", + "timepoint": 2443, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2443, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1028_0", + "timepoint": 1028, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1028, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; assistant,patient,disinfection; forceps,cotton,touching; assistant,antiseptic,holding " + } + ] + }, + { + "id": "MISS_3_6_141_0", + "timepoint": 141, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 141, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; head surgeon,microscope eye,looking; assistant,instrument table,closeto; head surgeon,herbal disk,removing; curette,patient,touching; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_666_0", + "timepoint": 666, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 666, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_668_0", + "timepoint": 668, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 668, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_158_0", + "timepoint": 158, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 158, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; scalpel,patient,touching; head surgeon,patient,cutting; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,scalpel,holding; head surgeon,operating table,closeto; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_1972_0", + "timepoint": 1972, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1972, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,instrument table,looking; assistant,antiseptic,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1518_0", + "timepoint": 1518, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1518, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_245_0", + "timepoint": 245, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 245, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; head surgeon,forceps,holding; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,curette,holding; anesthetist,health monitor,looking; head surgeon,herbal disk,removing; curette,patient,touching; forceps,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1104_0", + "timepoint": 1104, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1104, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; anesthetist,health monitor,looking; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1252_0", + "timepoint": 1252, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1252, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,patient,manipulating; assistant,operating table,closeto; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_542_0", + "timepoint": 542, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 542, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_180_0", + "timepoint": 180, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 180, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_713_0", + "timepoint": 713, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 713, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_512_0", + "timepoint": 512, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 512, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "MISS_3_6_877_0", + "timepoint": 877, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 877, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1559_0", + "timepoint": 1559, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1559, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_671_0", + "timepoint": 671, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 671, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,patient,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_502_0", + "timepoint": 502, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 502, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2148_0", + "timepoint": 2148, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2148, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; syringe,needle,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_801_0", + "timepoint": 801, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 801, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1334_0", + "timepoint": 1334, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1334, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,removing " + } + ] + }, + { + "id": "MISS_1_1_2022_0", + "timepoint": 2022, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2022, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,disinfection; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_594_0", + "timepoint": 594, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 594, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_787_0", + "timepoint": 787, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 787, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding; assistant,forceps,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2399_0", + "timepoint": 2399, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2399, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; needle,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; assistant,needle,holding; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1871_0", + "timepoint": 1871, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1871, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,needle,holding; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_898_0", + "timepoint": 898, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 898, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_484_0", + "timepoint": 484, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 484, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2862_0", + "timepoint": 2862, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2862, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,instrument table,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_765_0", + "timepoint": 765, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 765, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; head surgeon,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2272_0", + "timepoint": 2272, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2272, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; assistant,syringe,injecting; assistant,needle,holding; patient,operating table,lyingon; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_1_1_1461_0", + "timepoint": 1461, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1461, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,microscope controller,manipulating; assistant,operating table,closeto; anesthetist,patient,anaesthetising " + } + ] + }, + { + "id": "Ultrasound_3_4_1493_0", + "timepoint": 1493, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1493, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2574_0", + "timepoint": 2574, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2574, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_762_0", + "timepoint": 762, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 762, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope screen,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_312_0", + "timepoint": 312, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 312, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound gel,applying; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_27_0", + "timepoint": 27, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 27, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2772_0", + "timepoint": 2772, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2772, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; head surgeon,ultrasound probe,controlling; needle,patient,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_1146_0", + "timepoint": 1146, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1146, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating room,entering; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating room,entering " + } + ] + }, + { + "id": "MISS_3_6_746_0", + "timepoint": 746, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 746, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,instruments,disinfection; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; assistant,anesthetist,applying; anesthetist,health monitor,looking; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_222_0", + "timepoint": 222, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 222, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,instrument table,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1152_0", + "timepoint": 1152, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1152, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1648_0", + "timepoint": 1648, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1648, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,inserting; patient,operating table,lyingon; assistant,tissue mark,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_186_0", + "timepoint": 186, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 186, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,curette,holding; assistant,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,microscope eye,looking; curette,patient,touching; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_451_0", + "timepoint": 451, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 451, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1720_0", + "timepoint": 1720, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1720, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,tissue mark,looking; head surgeon,operating table,closeto; assistant,needle,inserting; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1568_0", + "timepoint": 1568, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1568, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,needle,inserting; ultrasound probe,patient,touching; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_966_0", + "timepoint": 966, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 966, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1389_0", + "timepoint": 1389, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1389, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,needle,removing; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1647_0", + "timepoint": 1647, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1647, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,inserting; assistant,tissue mark,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1804_0", + "timepoint": 1804, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1804, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,antiseptic,holding; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1832_0", + "timepoint": 1832, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1832, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,inserting; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2582_0", + "timepoint": 2582, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2582, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_820_0", + "timepoint": 820, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 820, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_980_0", + "timepoint": 980, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 980, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; cotton,patient,touching; forceps,cotton,touching; head surgeon,microscope eye,looking; assistant,patient,aspirating; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_400_0", + "timepoint": 400, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 400, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,forceps,holding; assistant,instrument table,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope eye,looking; curette,patient,touching; anesthetist,operating table,closeto; head surgeon,curette,holding; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "Ultrasound_5_6_1835_0", + "timepoint": 1835, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1835, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,syringe,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,needle,holding; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2881_0", + "timepoint": 2881, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2881, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,looking; assistant,instrument table,closeto; assistant,needle,holding; head surgeon,ultrasound machine,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1939_0", + "timepoint": 1939, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1939, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,ultrasound probe,looking; head surgeon,ultrasound screen,looking; assistant,syringe,holding; assistant,needle,inserting; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_145_0", + "timepoint": 145, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 145, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,manipulating; head surgeon,ultrasound machine,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3014_0", + "timepoint": 3014, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3014, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; tissue paper,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_2423_0", + "timepoint": 2423, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2423, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,needle,removing; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "MISS_2_2_96_0", + "timepoint": 96, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 96, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,scalpel,holding; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,patient,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2613_0", + "timepoint": 2613, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2613, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_3_4_1416_0", + "timepoint": 1416, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1416, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,ultrasound probe,holding; assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1025_0", + "timepoint": 1025, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1025, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1625_0", + "timepoint": 1625, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1625, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1748_0", + "timepoint": 1748, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1748, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,needle,holding; assistant,instrument table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_306_0", + "timepoint": 306, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 306, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; anesthetist,health monitor,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; head surgeon,scalpel,holding; assistant,operating table,closeto; assistant,patient,aspirating " + } + ] + }, + { + "id": "Ultrasound_4_5_124_0", + "timepoint": 124, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 124, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2055_0", + "timepoint": 2055, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2055, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,syringe,injecting; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_276_0", + "timepoint": 276, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 276, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,forceps,holding; forceps,patient,touching; curette,patient,touching; head surgeon,herbal disk,removing; assistant,operating table,closeto; anesthetist,health monitor,looking; head surgeon,curette,holding; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2120_0", + "timepoint": 2120, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2120, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; syringe,needle,touching; head surgeon,ultrasound screen,looking; assistant,needle,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_5_6_1942_0", + "timepoint": 1942, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1942, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,holding; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_3_4_2418_0", + "timepoint": 2418, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2418, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; syringe,needle,touching; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1042_0", + "timepoint": 1042, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1042, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,body marker,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_891_0", + "timepoint": 891, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 891, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1429_0", + "timepoint": 1429, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1429, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_714_0", + "timepoint": 714, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 714, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1699_0", + "timepoint": 1699, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1699, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,tissue mark,looking; head surgeon,vertebrae,scanning; assistant,needle,inserting; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_259_0", + "timepoint": 259, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 259, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; curette,patient,touching; head surgeon,microscope eye,looking; head surgeon,curette,holding; assistant,operating table,closeto; head surgeon,herbal disk,removing; forceps,patient,touching; anesthetist,health monitor,looking; head surgeon,forceps,holding; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_984_0", + "timepoint": 984, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 984, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2268_0", + "timepoint": 2268, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2268, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,removing; patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_321_0", + "timepoint": 321, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 321, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; forceps,patient,touching; head surgeon,forceps,holding; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing; curette,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2090_0", + "timepoint": 2090, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2090, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; syringe,needle,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_90_0", + "timepoint": 90, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 90, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,manipulating; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_557_0", + "timepoint": 557, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 557, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2993_0", + "timepoint": 2993, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2993, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_97_0", + "timepoint": 97, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 97, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1393_0", + "timepoint": 1393, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1393, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,removing; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_194_0", + "timepoint": 194, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 194, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2127_0", + "timepoint": 2127, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2127, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; needle,patient,touching; syringe,needle,touching; assistant,syringe,injecting; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1126_0", + "timepoint": 1126, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1126, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,body marker,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2409_0", + "timepoint": 2409, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2409, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,needle,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,injecting; assistant,needle,holding; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_292_0", + "timepoint": 292, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 292, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; assistant,operating table,closeto; forceps,patient,touching; head surgeon,curette,holding; head surgeon,microscope eye,looking; curette,patient,touching; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,health monitor,looking; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_76_0", + "timepoint": 76, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 76, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound machine,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_485_0", + "timepoint": 485, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 485, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_3046_0", + "timepoint": 3046, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3046, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,tissue paper,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_944_0", + "timepoint": 944, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 944, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,body marker,holding; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2462_0", + "timepoint": 2462, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2462, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; needle,patient,touching; head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2174_0", + "timepoint": 2174, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2174, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,ultrasound screen,looking; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,needle,holding; syringe,needle,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_3_4_612_0", + "timepoint": 612, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 612, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2507_0", + "timepoint": 2507, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2507, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; syringe,needle,touching; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2372_0", + "timepoint": 2372, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2372, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1010_0", + "timepoint": 1010, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1010, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_1245_0", + "timepoint": 1245, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1245, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,operating table,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2912_0", + "timepoint": 2912, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2912, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_3280_0", + "timepoint": 3280, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3280, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,tissue paper,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2602_0", + "timepoint": 2602, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2602, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; assistant,syringe,injecting; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_776_0", + "timepoint": 776, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 776, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1695_0", + "timepoint": 1695, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1695, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,tissue mark,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_468_0", + "timepoint": 468, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 468, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_3095_0", + "timepoint": 3095, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3095, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2933_0", + "timepoint": 2933, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2933, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; assistant,operating table,closeto; tissue paper,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1298_0", + "timepoint": 1298, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1298, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1411_0", + "timepoint": 1411, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1411, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope controller,holding; patient,operating table,lyingon; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_584_0", + "timepoint": 584, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 584, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1210_0", + "timepoint": 1210, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1210, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; circulator,ultrasound machine,manipulating; assistant,antiseptic,holding; assistant,operating table,closeto; patient,operating table,lyingon; circulator,ultrasound machine,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1416_0", + "timepoint": 1416, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1416, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,needle,removing; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_897_0", + "timepoint": 897, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 897, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,scissors,holding; circulator,ultrasound machine,closeto; circulator,ultrasound machine,manipulating; patient,operating table,lyingon; assistant,operating table,closeto; scissors,cotton,touching; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1706_0", + "timepoint": 1706, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1706, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,syringe,holding; head surgeon,vertebrae,scanning; assistant,instrument table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_2208_0", + "timepoint": 2208, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2208, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; syringe,needle,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_391_0", + "timepoint": 391, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 391, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,herbal disk,removing; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,curette,holding; assistant,forceps,holding; head surgeon,forceps,holding; curette,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_987_0", + "timepoint": 987, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 987, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; cotton,patient,touching; assistant,patient,disinfection; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3099_0", + "timepoint": 3099, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3099, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2488_0", + "timepoint": 2488, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2488, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,looking; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2249_0", + "timepoint": 2249, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2249, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching; syringe,needle,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_976_0", + "timepoint": 976, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 976, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,antiseptic,holding; assistant,operating table,closeto; assistant,patient,disinfection; cotton,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_1_1_505_0", + "timepoint": 505, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 505, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_313_0", + "timepoint": 313, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 313, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_1345_0", + "timepoint": 1345, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1345, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; needle,patient,touching; assistant,needle,removing; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_317_0", + "timepoint": 317, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 317, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,forceps,holding; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; forceps,patient,touching; head surgeon,curette,holding; head surgeon,operating table,closeto; curette,patient,touching; anesthetist,health monitor,looking; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_1113_0", + "timepoint": 1113, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1113, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,manipulating; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding; assistant,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; circulator,ultrasound machine,closeto " + } + ] + }, + { + "id": "MISS_1_1_2396_0", + "timepoint": 2396, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2396, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1072_0", + "timepoint": 1072, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1072, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1100_0", + "timepoint": 1100, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1100, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,body marker,holding " + } + ] + }, + { + "id": "MISS_1_1_1542_0", + "timepoint": 1542, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1542, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,microscope controller,manipulating; anesthetist,patient,anaesthetising; head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2250_0", + "timepoint": 2250, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2250, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; syringe,needle,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,needle,holding; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1156_0", + "timepoint": 1156, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1156, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2276_0", + "timepoint": 2276, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2276, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2293_0", + "timepoint": 2293, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2293, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_859_0", + "timepoint": 859, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 859, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,applying; body marker,patient,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2912_0", + "timepoint": 2912, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2912, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "MISS_3_6_586_0", + "timepoint": 586, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 586, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; curette,patient,touching; head surgeon,microscope eye,looking; patient,operating table,lyingon; head surgeon,herbal disk,removing; scalpel,patient,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2408_0", + "timepoint": 2408, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2408, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1047_0", + "timepoint": 1047, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1047, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; patient,operating table,lyingon; assistant,forceps,holding; anesthetist,operating table,closeto; assistant,patient,aspirating; cotton,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,microscope eye,looking; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_3110_0", + "timepoint": 3110, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3110, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_920_0", + "timepoint": 920, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 920, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1180_0", + "timepoint": 1180, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1180, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating room,entering; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2401_0", + "timepoint": 2401, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2401, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_596_0", + "timepoint": 596, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 596, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2584_0", + "timepoint": 2584, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2584, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,tissue paper,applying; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_910_0", + "timepoint": 910, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 910, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,body marker,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2964_0", + "timepoint": 2964, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2964, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; tissue paper,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_591_0", + "timepoint": 591, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 591, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_228_0", + "timepoint": 228, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 228, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; head surgeon,herbal disk,removing; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; assistant,operating table,closeto; head surgeon,forceps,holding; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2879_0", + "timepoint": 2879, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2879, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,instrument table,looking; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,ultrasound machine,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2266_0", + "timepoint": 2266, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2266, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; anesthetist,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1622_0", + "timepoint": 1622, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1622, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3093_0", + "timepoint": 3093, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3093, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1327_0", + "timepoint": 1327, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1327, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; forceps,cotton,touching; assistant,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_41_0", + "timepoint": 41, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 41, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,instrument table,looking; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_2780_0", + "timepoint": 2780, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2780, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,needle,removing " + } + ] + }, + { + "id": "Ultrasound_3_4_2246_0", + "timepoint": 2246, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2246, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; patient,operating table,lyingon; assistant,needle,holding; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2954_0", + "timepoint": 2954, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2954, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,removing; head surgeon,patient,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; assistant,syringe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_148_0", + "timepoint": 148, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 148, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2592_0", + "timepoint": 2592, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2592, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1017_0", + "timepoint": 1017, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1017, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; assistant,patient,disinfection; assistant,antiseptic,holding; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2659_0", + "timepoint": 2659, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2659, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,needle,holding; needle,patient,touching; ultrasound probe,patient,touching; syringe,needle,touching; assistant,operating table,closeto; assistant,syringe,injecting; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_225_0", + "timepoint": 225, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 225, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_49_0", + "timepoint": 49, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 49, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1559_0", + "timepoint": 1559, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1559, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; assistant,needle,inserting; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_629_0", + "timepoint": 629, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 629, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,manipulating; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1663_0", + "timepoint": 1663, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1663, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_355_0", + "timepoint": 355, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 355, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,operating table,looking; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2533_0", + "timepoint": 2533, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2533, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2105_0", + "timepoint": 2105, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2105, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,disinfection; forceps,cotton,touching; cotton,patient,touching; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_480_0", + "timepoint": 480, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 480, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_378_0", + "timepoint": 378, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 378, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_669_0", + "timepoint": 669, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 669, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2039_0", + "timepoint": 2039, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2039, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; needle,patient,touching; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,needle,inserting; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2603_0", + "timepoint": 2603, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2603, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_3_4_1413_0", + "timepoint": 1413, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1413, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,body marker,holding; head surgeon,ultrasound probe,holding; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_856_0", + "timepoint": 856, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 856, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_590_0", + "timepoint": 590, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 590, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1399_0", + "timepoint": 1399, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1399, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,removing; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_486_0", + "timepoint": 486, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 486, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_2098_0", + "timepoint": 2098, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2098, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; needle,patient,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_2_2_1157_0", + "timepoint": 1157, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1157, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,manipulating; head surgeon,operating table,closeto; assistant,antiseptic,holding; assistant,operating table,closeto; circulator,ultrasound machine,closeto; assistant,forceps,holding; patient,operating table,lyingon; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1427_0", + "timepoint": 1427, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1427, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,removing; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_694_0", + "timepoint": 694, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 694, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2525_0", + "timepoint": 2525, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2525, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,needle,holding; assistant,syringe,injecting; syringe,needle,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1181_0", + "timepoint": 1181, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1181, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_444_0", + "timepoint": 444, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 444, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1644_0", + "timepoint": 1644, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1644, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1394_0", + "timepoint": 1394, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1394, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,needle,removing " + } + ] + }, + { + "id": "Ultrasound_3_4_2365_0", + "timepoint": 2365, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2365, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching; assistant,syringe,injecting; assistant,needle,holding; syringe,needle,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_1965_0", + "timepoint": 1965, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1965, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,patient,disinfection; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_348_0", + "timepoint": 348, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 348, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; circulator,operating table,looking; patient,operating table,closeto; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1118_0", + "timepoint": 1118, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1118, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,inserting; assistant,ultrasound probe,looking; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_624_0", + "timepoint": 624, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 624, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_892_0", + "timepoint": 892, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 892, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,microscope,manipulating " + } + ] + }, + { + "id": "MISS_3_6_103_0", + "timepoint": 103, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 103, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,herbal disk,removing; anesthetist,operating table,closeto; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2971_0", + "timepoint": 2971, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2971, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_79_0", + "timepoint": 79, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 79, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,scalpel,holding; patient,operating table,lyingon; head surgeon,patient,manipulating; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1662_0", + "timepoint": 1662, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1662, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_893_0", + "timepoint": 893, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 893, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1087_0", + "timepoint": 1087, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1087, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,body marker,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_564_0", + "timepoint": 564, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 564, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; curette,patient,touching; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; scalpel,patient,touching; head surgeon,herbal disk,removing; head surgeon,scalpel,holding; head surgeon,microscope eye,looking; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1594_0", + "timepoint": 1594, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1594, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_314_0", + "timepoint": 314, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 314, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,patient,touching; head surgeon,herbal disk,removing; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,operating table,closeto; head surgeon,curette,holding; head surgeon,forceps,holding; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1239_0", + "timepoint": 1239, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1239, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2356_0", + "timepoint": 2356, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2356, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_985_0", + "timepoint": 985, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 985, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,microscope,manipulating; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1966_0", + "timepoint": 1966, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1966, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; needle,patient,touching; assistant,ultrasound probe,looking; assistant,needle,inserting; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,syringe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_812_0", + "timepoint": 812, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 812, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1800_0", + "timepoint": 1800, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1800, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,antiseptic,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2188_0", + "timepoint": 2188, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2188, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,needle,holding; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2766_0", + "timepoint": 2766, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2766, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; head surgeon,vertebrae,scanning; head surgeon,needle,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_935_0", + "timepoint": 935, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 935, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,patient,aspirating; assistant,forceps,holding; cotton,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,microscope eye,looking; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_531_0", + "timepoint": 531, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 531, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "MISS_1_1_2504_0", + "timepoint": 2504, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2504, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,patient,manipulating; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_717_0", + "timepoint": 717, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 717, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,looking; head surgeon,microscope eye,looking; assistant,instrument table,closeto; assistant,curette,holding; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2016_0", + "timepoint": 2016, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2016, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; needle,patient,touching; head surgeon,needle,holding; assistant,instrument table,closeto; assistant,syringe,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3137_0", + "timepoint": 3137, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3137, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,applying " + } + ] + }, + { + "id": "MISS_1_1_2270_0", + "timepoint": 2270, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2270, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1867_0", + "timepoint": 1867, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1867, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,applying; assistant,patient,disinfection; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_316_0", + "timepoint": 316, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 316, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; assistant,operating table,closeto; assistant,forceps,holding; head surgeon,scalpel,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,patient,aspirating; anesthetist,operating table,closeto; anesthetist,health monitor,looking; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_34_0", + "timepoint": 34, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 34, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_292_0", + "timepoint": 292, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 292, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,scalpel,holding; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_255_0", + "timepoint": 255, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 255, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_262_0", + "timepoint": 262, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 262, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_807_0", + "timepoint": 807, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 807, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,microscope,closeto; anesthetist,operating table,closeto; circulator,microscope,manipulating; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "MISS_3_6_202_0", + "timepoint": 202, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 202, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,forceps,holding; patient,operating table,lyingon; curette,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_1202_0", + "timepoint": 1202, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1202, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1235_0", + "timepoint": 1235, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1235, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1157_0", + "timepoint": 1157, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1157, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating room,entering " + } + ] + }, + { + "id": "MISS_1_1_857_0", + "timepoint": 857, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 857, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,microscope,manipulating; patient,operating table,lyingon; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_3112_0", + "timepoint": 3112, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3112, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_34_0", + "timepoint": 34, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 34, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1992_0", + "timepoint": 1992, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1992, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_5_6_229_0", + "timepoint": 229, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 229, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2562_0", + "timepoint": 2562, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2562, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon; assistant,operating table,closeto; assistant,syringe,injecting; syringe,needle,touching; needle,patient,touching; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2465_0", + "timepoint": 2465, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2465, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,patient,manipulating; assistant,patient,manipulating " + } + ] + }, + { + "id": "MISS_1_1_2061_0", + "timepoint": 2061, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2061, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,patient,disinfection; assistant,instrument table,closeto; forceps,cotton,touching; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_505_0", + "timepoint": 505, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 505, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_258_0", + "timepoint": 258, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 258, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating room,entering; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_279_0", + "timepoint": 279, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 279, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,scalpel,holding; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,health monitor,looking; forceps,cotton,touching; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_405_0", + "timepoint": 405, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 405, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,touching; circulator,operating table,closeto; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2379_0", + "timepoint": 2379, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2379, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,holding; needle,patient,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_278_0", + "timepoint": 278, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 278, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_24_0", + "timepoint": 24, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 24, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_766_0", + "timepoint": 766, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 766, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_191_0", + "timepoint": 191, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 191, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,curette,holding; curette,patient,touching; head surgeon,herbal disk,removing; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1449_0", + "timepoint": 1449, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1449, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1520_0", + "timepoint": 1520, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1520, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_50_0", + "timepoint": 50, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 50, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,scalpel,looking; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_1033_0", + "timepoint": 1033, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1033, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; assistant,forceps,holding; patient,operating table,lyingon; forceps,cotton,touching; assistant,operating table,closeto; cotton,patient,touching; head surgeon,operating table,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "MISS_2_2_614_0", + "timepoint": 614, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 614, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,controlling; anesthetist,operating table,closeto; head surgeon,microscope screen,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,operating table,closeto; assistant,microscope screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1520_0", + "timepoint": 1520, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1520, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope controller,manipulating; patient,operating table,lyingon; anesthetist,patient,anaesthetising " + } + ] + }, + { + "id": "Ultrasound_3_4_2014_0", + "timepoint": 2014, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2014, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1466_0", + "timepoint": 1466, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1466, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,needle,holding; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_45_0", + "timepoint": 45, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 45, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_3_6_1000_0", + "timepoint": 1000, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1000, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; anesthetist,operating table,closeto; assistant,patient,aspirating; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; forceps,cotton,touching; cotton,patient,touching; assistant,forceps,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_162_0", + "timepoint": 162, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 162, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound machine,looking; assistant,operating table,closeto; assistant,patient,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_410_0", + "timepoint": 410, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 410, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2839_0", + "timepoint": 2839, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2839, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_380_0", + "timepoint": 380, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 380, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering; circulator,operating table,manipulating; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_3196_0", + "timepoint": 3196, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3196, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_577_0", + "timepoint": 577, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 577, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_551_0", + "timepoint": 551, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 551, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_492_0", + "timepoint": 492, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 492, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_199_0", + "timepoint": 199, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 199, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound gel,applying; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1736_0", + "timepoint": 1736, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1736, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,inserting; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,tissue mark,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_451_0", + "timepoint": 451, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 451, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1435_0", + "timepoint": 1435, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1435, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_3098_0", + "timepoint": 3098, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3098, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "MISS_2_2_638_0", + "timepoint": 638, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 638, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope screen,looking; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,controlling; patient,operating table,lyingon; assistant,microscope screen,looking " + } + ] + }, + { + "id": "MISS_3_6_374_0", + "timepoint": 374, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 374, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; curette,patient,touching; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,microscope eye,looking; forceps,patient,touching; head surgeon,herbal disk,removing; head surgeon,curette,holding; head surgeon,forceps,holding; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1046_0", + "timepoint": 1046, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1046, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3278_0", + "timepoint": 3278, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3278, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,tissue paper,applying; tissue paper,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_178_0", + "timepoint": 178, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 178, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_338_0", + "timepoint": 338, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 338, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1513_0", + "timepoint": 1513, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1513, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_2025_0", + "timepoint": 2025, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2025, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,patient,disinfection; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1034_0", + "timepoint": 1034, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1034, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; cotton,patient,touching; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,patient,aspirating; assistant,forceps,holding; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,health monitor,looking; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2670_0", + "timepoint": 2670, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2670, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; assistant,needle,holding; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,vertebrae,scanning; syringe,needle,touching; patient,operating table,lyingon; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_3_6_378_0", + "timepoint": 378, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 378, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; head surgeon,curette,holding; head surgeon,microscope eye,looking; patient,operating table,lyingon; curette,patient,touching; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,forceps,holding; anesthetist,anesthesia equipment,controlling; forceps,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_782_0", + "timepoint": 782, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 782, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; circulator,microscope,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_403_0", + "timepoint": 403, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 403, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,touching; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_436_0", + "timepoint": 436, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 436, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_14_0", + "timepoint": 14, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 14, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering " + } + ] + }, + { + "id": "MISS_3_6_343_0", + "timepoint": 343, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 343, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; forceps,patient,touching; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; curette,patient,touching; head surgeon,forceps,holding; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_977_0", + "timepoint": 977, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 977, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,patient,aspirating; forceps,cotton,touching; anesthetist,operating table,closeto; cotton,patient,touching; head surgeon,microscope eye,looking; assistant,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1238_0", + "timepoint": 1238, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1238, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_801_0", + "timepoint": 801, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 801, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; assistant,antiseptic,holding; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2229_0", + "timepoint": 2229, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2229, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; syringe,needle,touching; assistant,needle,holding; needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1774_0", + "timepoint": 1774, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1774, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; assistant,syringe,holding; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1760_0", + "timepoint": 1760, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1760, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1786_0", + "timepoint": 1786, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1786, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; assistant,antiseptic,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2797_0", + "timepoint": 2797, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2797, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; patient,operating table,lyingon; assistant,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; needle,patient,touching; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_1125_0", + "timepoint": 1125, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1125, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,health monitor,looking; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2664_0", + "timepoint": 2664, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2664, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,tissue paper,holding; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1169_0", + "timepoint": 1169, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1169, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,forceps,holding; circulator,ultrasound machine,closeto; assistant,antiseptic,holding; circulator,ultrasound machine,manipulating; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_485_0", + "timepoint": 485, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 485, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; anesthetist,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_3005_0", + "timepoint": 3005, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3005, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; tissue paper,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2221_0", + "timepoint": 2221, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2221, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching; anesthetist,operating table,closeto; assistant,patient,disinfection; cotton,patient,touching; assistant,forceps,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1718_0", + "timepoint": 1718, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1718, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; anesthetist,operating table,closeto; assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2542_0", + "timepoint": 2542, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2542, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; needle,patient,touching; syringe,needle,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_1091_0", + "timepoint": 1091, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1091, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_1572_0", + "timepoint": 1572, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1572, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_97_0", + "timepoint": 97, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 97, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,antiseptic,looking; circulator,instrument table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_1_1_1912_0", + "timepoint": 1912, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1912, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,antiseptic,applying; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2433_0", + "timepoint": 2433, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2433, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound machine,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; assistant,needle,removing; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_1_1_667_0", + "timepoint": 667, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 667, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2571_0", + "timepoint": 2571, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2571, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,needle,holding; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; syringe,needle,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2486_0", + "timepoint": 2486, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2486, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,instrument table,looking; head surgeon,ultrasound machine,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1055_0", + "timepoint": 1055, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1055, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,patient,aspirating; head surgeon,microscope eye,looking; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; anesthetist,health monitor,looking; cotton,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3007_0", + "timepoint": 3007, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3007, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_90_0", + "timepoint": 90, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 90, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1232_0", + "timepoint": 1232, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1232, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,needle,inserting; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_488_0", + "timepoint": 488, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 488, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,scalpel,holding; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1149_0", + "timepoint": 1149, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1149, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_621_0", + "timepoint": 621, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 621, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_769_0", + "timepoint": 769, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 769, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,health monitor,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1081_0", + "timepoint": 1081, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1081, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_2315_0", + "timepoint": 2315, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2315, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_995_0", + "timepoint": 995, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 995, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,manipulating; anesthetist,microscope,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1026_0", + "timepoint": 1026, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1026, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; assistant,antiseptic,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; cotton,patient,touching; assistant,forceps,holding; patient,operating table,lyingon; forceps,cotton,touching " + } + ] + }, + { + "id": "MISS_1_1_2327_0", + "timepoint": 2327, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2327, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1084_0", + "timepoint": 1084, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1084, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; assistant,operating table,closeto; forceps,cotton,touching; circulator,ultrasound machine,looking; head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2155_0", + "timepoint": 2155, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2155, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; assistant,operating table,closeto; assistant,needle,holding; syringe,needle,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_58_0", + "timepoint": 58, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 58, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,patient,manipulating; assistant,scalpel,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2044_0", + "timepoint": 2044, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2044, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; patient,operating table,lyingon; syringe,needle,touching; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_176_0", + "timepoint": 176, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 176, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound machine,looking; patient,operating table,lyingon; assistant,operating table,closeto; assistant,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1721_0", + "timepoint": 1721, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1721, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,needle,inserting; needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_198_0", + "timepoint": 198, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 198, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling; circulator,scalpel,looking " + } + ] + }, + { + "id": "MISS_1_1_2218_0", + "timepoint": 2218, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2218, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,forceps,holding; assistant,patient,disinfection; cotton,patient,touching; forceps,cotton,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2011_0", + "timepoint": 2011, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2011, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,needle,inserting; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_1218_0", + "timepoint": 1218, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1218, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,closeto; assistant,forceps,holding; assistant,antiseptic,holding; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; circulator,ultrasound machine,manipulating; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1403_0", + "timepoint": 1403, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1403, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1756_0", + "timepoint": 1756, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1756, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,syringe,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_2259_0", + "timepoint": 2259, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2259, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,disinfection; assistant,forceps,holding; head surgeon,operating table,closeto; forceps,cotton,touching; cotton,patient,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_800_0", + "timepoint": 800, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 800, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; circulator,microscope,manipulating; anesthetist,operating table,closeto; circulator,microscope,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1136_0", + "timepoint": 1136, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1136, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2443_0", + "timepoint": 2443, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2443, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,manipulating; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_971_0", + "timepoint": 971, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 971, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1127_0", + "timepoint": 1127, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1127, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1567_0", + "timepoint": 1567, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1567, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,needle,inserting; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,ultrasound probe,looking; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1905_0", + "timepoint": 1905, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1905, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,needle,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2463_0", + "timepoint": 2463, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2463, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; anesthetist,operating table,closeto; head surgeon,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1900_0", + "timepoint": 1900, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1900, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1451_0", + "timepoint": 1451, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1451, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2178_0", + "timepoint": 2178, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2178, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; cotton,patient,touching; anesthetist,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "MISS_1_1_1532_0", + "timepoint": 1532, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1532, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope controller,manipulating; assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,patient,anaesthetising " + } + ] + }, + { + "id": "MISS_2_2_352_0", + "timepoint": 352, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 352, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; cotton,patient,touching; forceps,cotton,touching; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,scalpel,holding; assistant,forceps,holding; patient,operating table,lyingon; assistant,patient,aspirating " + } + ] + }, + { + "id": "Ultrasound_3_4_73_0", + "timepoint": 73, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 73, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_184_0", + "timepoint": 184, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 184, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,ultrasound gel,applying; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_650_0", + "timepoint": 650, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 650, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_337_0", + "timepoint": 337, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 337, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,scalpel,holding; anesthetist,health monitor,looking; head surgeon,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,patient,aspirating; cotton,patient,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1691_0", + "timepoint": 1691, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1691, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1248_0", + "timepoint": 1248, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1248, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2803_0", + "timepoint": 2803, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2803, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; assistant,needle,removing; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1912_0", + "timepoint": 1912, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1912, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,syringe,holding; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,needle,applying; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_793_0", + "timepoint": 793, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 793, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,microscope,manipulating; circulator,microscope,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2785_0", + "timepoint": 2785, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2785, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; head surgeon,needle,holding; patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1352_0", + "timepoint": 1352, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1352, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,body marker,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1932_0", + "timepoint": 1932, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1932, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,inserting; assistant,syringe,holding; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1552_0", + "timepoint": 1552, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1552, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1558_0", + "timepoint": 1558, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1558, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,needle,inserting; assistant,ultrasound probe,looking; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_94_0", + "timepoint": 94, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 94, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,body marker,looking; circulator,instrument table,closeto; circulator,instruments,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_215_0", + "timepoint": 215, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 215, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_3_6_868_0", + "timepoint": 868, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 868, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3290_0", + "timepoint": 3290, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3290, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,tissue paper,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2032_0", + "timepoint": 2032, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2032, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; syringe,needle,touching; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_472_0", + "timepoint": 472, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 472, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1195_0", + "timepoint": 1195, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1195, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_822_0", + "timepoint": 822, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 822, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_120_0", + "timepoint": 120, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 120, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1280_0", + "timepoint": 1280, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1280, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,antiseptic,holding; assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1680_0", + "timepoint": 1680, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1680, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue mark,looking; assistant,needle,inserting; needle,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1344_0", + "timepoint": 1344, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1344, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2056_0", + "timepoint": 2056, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2056, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,patient,disinfection; assistant,forceps,holding; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1099_0", + "timepoint": 1099, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1099, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; circulator,ultrasound machine,looking; assistant,forceps,holding; assistant,antiseptic,holding; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; circulator,ultrasound machine,closeto " + } + ] + }, + { + "id": "MISS_1_1_1785_0", + "timepoint": 1785, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1785, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,antiseptic,holding; assistant,instrument table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1465_0", + "timepoint": 1465, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1465, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_611_0", + "timepoint": 611, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 611, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_3067_0", + "timepoint": 3067, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3067, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1789_0", + "timepoint": 1789, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1789, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,needle,holding; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,syringe,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_602_0", + "timepoint": 602, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 602, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_1925_0", + "timepoint": 1925, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1925, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,needle,holding; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_1_1_33_0", + "timepoint": 33, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 33, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_3_6_535_0", + "timepoint": 535, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 535, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; scalpel,patient,touching; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,scalpel,holding; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1987_0", + "timepoint": 1987, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1987, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; assistant,operating table,closeto; syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1600_0", + "timepoint": 1600, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1600, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_720_0", + "timepoint": 720, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 720, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "MISS_1_1_409_0", + "timepoint": 409, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 409, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_655_0", + "timepoint": 655, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 655, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_864_0", + "timepoint": 864, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 864, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_415_0", + "timepoint": 415, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 415, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1830_0", + "timepoint": 1830, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1830, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,syringe,holding; assistant,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2557_0", + "timepoint": 2557, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2557, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,needle,holding; assistant,operating table,closeto; syringe,needle,touching; assistant,syringe,injecting; assistant,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1288_0", + "timepoint": 1288, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1288, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1940_0", + "timepoint": 1940, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1940, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,needle,inserting; assistant,syringe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1654_0", + "timepoint": 1654, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1654, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound screen,looking; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_427_0", + "timepoint": 427, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 427, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3153_0", + "timepoint": 3153, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3153, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; patient,operating table,lyingon; assistant,tissue paper,applying; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2824_0", + "timepoint": 2824, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2824, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; head surgeon,ultrasound machine,looking; assistant,operating table,closeto; assistant,needle,removing " + } + ] + }, + { + "id": "Ultrasound_4_5_2610_0", + "timepoint": 2610, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2610, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_2330_0", + "timepoint": 2330, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2330, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2142_0", + "timepoint": 2142, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2142, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon; syringe,needle,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1477_0", + "timepoint": 1477, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1477, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_612_0", + "timepoint": 612, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 612, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_1236_0", + "timepoint": 1236, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1236, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; forceps,cotton,touching; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_327_0", + "timepoint": 327, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 327, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,manipulating; patient,operating room,entering; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1843_0", + "timepoint": 1843, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1843, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; needle,patient,touching; syringe,needle,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,needle,holding; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_2_2_1207_0", + "timepoint": 1207, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1207, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,closeto; forceps,cotton,touching; assistant,operating table,closeto; circulator,ultrasound machine,manipulating; assistant,forceps,holding; assistant,antiseptic,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2223_0", + "timepoint": 2223, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2223, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,needle,holding; syringe,needle,touching; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_478_0", + "timepoint": 478, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 478, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_2384_0", + "timepoint": 2384, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2384, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,syringe,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_104_0", + "timepoint": 104, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 104, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_535_0", + "timepoint": 535, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 535, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1173_0", + "timepoint": 1173, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1173, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating room,entering; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_738_0", + "timepoint": 738, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 738, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2590_0", + "timepoint": 2590, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2590, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,tissue paper,applying; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2480_0", + "timepoint": 2480, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2480, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning; head surgeon,needle,holding; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2609_0", + "timepoint": 2609, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2609, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching; syringe,needle,touching; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1036_0", + "timepoint": 1036, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1036, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,body marker,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2357_0", + "timepoint": 2357, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2357, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,syringe,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_797_0", + "timepoint": 797, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 797, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2401_0", + "timepoint": 2401, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2401, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; assistant,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_1_1_112_0", + "timepoint": 112, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 112, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2675_0", + "timepoint": 2675, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2675, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; needle,patient,touching; assistant,needle,holding; syringe,needle,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_2107_0", + "timepoint": 2107, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2107, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; forceps,cotton,touching; assistant,patient,disinfection; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; cotton,patient,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_240_0", + "timepoint": 240, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 240, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "MISS_3_6_368_0", + "timepoint": 368, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 368, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; head surgeon,forceps,holding; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; forceps,patient,touching; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_758_0", + "timepoint": 758, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 758, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2158_0", + "timepoint": 2158, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2158, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto; assistant,syringe,injecting; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_706_0", + "timepoint": 706, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 706, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1938_0", + "timepoint": 1938, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1938, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,holding; assistant,operating table,closeto; assistant,needle,inserting; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_355_0", + "timepoint": 355, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 355, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,forceps,holding; head surgeon,herbal disk,removing; anesthetist,operating table,closeto; curette,patient,touching; head surgeon,curette,holding; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; forceps,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_253_0", + "timepoint": 253, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 253, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,patient,looking " + } + ] + }, + { + "id": "MISS_1_1_91_0", + "timepoint": 91, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 91, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,antiseptic,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1262_0", + "timepoint": 1262, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1262, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding; anesthetist,patient,manipulating; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2470_0", + "timepoint": 2470, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2470, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,operating table,closeto; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; syringe,needle,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_532_0", + "timepoint": 532, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 532, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_283_0", + "timepoint": 283, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 283, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,patient,touching; head surgeon,herbal disk,removing; head surgeon,curette,holding; curette,patient,touching; head surgeon,forceps,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_962_0", + "timepoint": 962, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 962, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1406_0", + "timepoint": 1406, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1406, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_618_0", + "timepoint": 618, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 618, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2118_0", + "timepoint": 2118, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2118, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; syringe,needle,touching; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,needle,holding; needle,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_1091_0", + "timepoint": 1091, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1091, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; head surgeon,microscope eye,looking; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,health monitor,looking; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_463_0", + "timepoint": 463, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 463, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_389_0", + "timepoint": 389, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 389, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_43_0", + "timepoint": 43, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 43, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_1_1_333_0", + "timepoint": 333, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 333, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,operating table,manipulating; anesthetist,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1048_0", + "timepoint": 1048, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1048, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_515_0", + "timepoint": 515, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 515, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1684_0", + "timepoint": 1684, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1684, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_1365_0", + "timepoint": 1365, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1365, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,instrument table,looking; head surgeon,operating table,closeto; circulator,operating table,looking; patient,operating table,lyingon; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1384_0", + "timepoint": 1384, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1384, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,body marker,holding; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2629_0", + "timepoint": 2629, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2629, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; needle,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1139_0", + "timepoint": 1139, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1139, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,body marker,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2344_0", + "timepoint": 2344, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2344, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,syringe,holding; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_543_0", + "timepoint": 543, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 543, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1035_0", + "timepoint": 1035, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1035, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,body marker,holding; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1146_0", + "timepoint": 1146, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1146, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; circulator,ultrasound machine,manipulating; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,antiseptic,holding; circulator,ultrasound machine,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_420_0", + "timepoint": 420, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 420, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2092_0", + "timepoint": 2092, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2092, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,needle,holding; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_201_0", + "timepoint": 201, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 201, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,ultrasound gel,looking; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1271_0", + "timepoint": 1271, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1271, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,needle,inserting; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1285_0", + "timepoint": 1285, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1285, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; assistant,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_1_1_1898_0", + "timepoint": 1898, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1898, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,applying; anesthetist,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_3_4_403_0", + "timepoint": 403, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 403, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_885_0", + "timepoint": 885, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 885, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,applying; assistant,patient,disinfection; assistant,operating table,closeto; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_345_0", + "timepoint": 345, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 345, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,applying " + } + ] + }, + { + "id": "Ultrasound_3_4_706_0", + "timepoint": 706, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 706, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2047_0", + "timepoint": 2047, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2047, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_2462_0", + "timepoint": 2462, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2462, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_2680_0", + "timepoint": 2680, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2680, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,needle,holding; assistant,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_48_0", + "timepoint": 48, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 48, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto; circulator,forceps,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_3301_0", + "timepoint": 3301, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3301, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,bin,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,tissue paper,holding " + } + ] + }, + { + "id": "MISS_1_1_886_0", + "timepoint": 886, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 886, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "MISS_1_1_2322_0", + "timepoint": 2322, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2322, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1431_0", + "timepoint": 1431, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1431, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_277_0", + "timepoint": 277, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 277, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; curette,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; anesthetist,health monitor,looking; head surgeon,curette,holding; head surgeon,forceps,holding; forceps,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_3027_0", + "timepoint": 3027, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3027, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; tissue paper,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2805_0", + "timepoint": 2805, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2805, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,needle,removing; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1184_0", + "timepoint": 1184, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1184, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,inserting; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_260_0", + "timepoint": 260, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 260, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,ultrasound gel,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_934_0", + "timepoint": 934, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 934, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2434_0", + "timepoint": 2434, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2434, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_243_0", + "timepoint": 243, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 243, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1905_0", + "timepoint": 1905, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1905, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,patient,disinfection; assistant,operating table,closeto; assistant,antiseptic,applying; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2850_0", + "timepoint": 2850, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2850, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,syringe,removing; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound screen,looking; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1817_0", + "timepoint": 1817, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1817, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_3171_0", + "timepoint": 3171, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3171, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,applying " + } + ] + }, + { + "id": "MISS_1_1_2184_0", + "timepoint": 2184, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2184, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,patient,disinfection; assistant,operating table,closeto; cotton,patient,touching; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_695_0", + "timepoint": 695, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 695, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_665_0", + "timepoint": 665, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 665, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_822_0", + "timepoint": 822, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 822, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,body marker,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_127_0", + "timepoint": 127, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 127, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2350_0", + "timepoint": 2350, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2350, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_898_0", + "timepoint": 898, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 898, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1314_0", + "timepoint": 1314, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1314, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound probe,looking; assistant,body marker,applying; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; body marker,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_776_0", + "timepoint": 776, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 776, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,forceps,holding; patient,operating table,lyingon; forceps,cotton,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1306_0", + "timepoint": 1306, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1306, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,antiseptic,holding; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_212_0", + "timepoint": 212, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 212, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound gel,looking; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_311_0", + "timepoint": 311, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 311, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "MISS_1_1_2232_0", + "timepoint": 2232, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2232, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; assistant,patient,disinfection; cotton,patient,touching; assistant,forceps,holding; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_249_0", + "timepoint": 249, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 249, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,forceps,holding; scalpel,patient,touching; head surgeon,scalpel,holding; anesthetist,health monitor,looking; head surgeon,patient,cutting; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_291_0", + "timepoint": 291, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 291, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2846_0", + "timepoint": 2846, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2846, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,holding; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3025_0", + "timepoint": 3025, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3025, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,tissue paper,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_191_0", + "timepoint": 191, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 191, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,applying; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_3_6_315_0", + "timepoint": 315, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 315, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,health monitor,looking; head surgeon,curette,holding; head surgeon,herbal disk,removing; anesthetist,operating table,closeto; head surgeon,forceps,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; forceps,patient,touching; head surgeon,microscope eye,looking; curette,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2886_0", + "timepoint": 2886, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2886, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,syringe,removing; syringe,needle,touching; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,needle,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_835_0", + "timepoint": 835, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 835, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1690_0", + "timepoint": 1690, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1690, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; assistant,needle,inserting; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_528_0", + "timepoint": 528, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 528, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_619_0", + "timepoint": 619, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 619, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1177_0", + "timepoint": 1177, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1177, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_12_0", + "timepoint": 12, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 12, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_2834_0", + "timepoint": 2834, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2834, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; assistant,syringe,removing; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_959_0", + "timepoint": 959, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 959, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_1276_0", + "timepoint": 1276, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1276, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; forceps,cotton,touching; assistant,antiseptic,holding; assistant,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1763_0", + "timepoint": 1763, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1763, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,ultrasound screen,looking; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1136_0", + "timepoint": 1136, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1136, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; assistant,needle,inserting; needle,patient,touching; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_147_0", + "timepoint": 147, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 147, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_82_0", + "timepoint": 82, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 82, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_337_0", + "timepoint": 337, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 337, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_1880_0", + "timepoint": 1880, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1880, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2558_0", + "timepoint": 2558, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2558, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,syringe,injecting; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2387_0", + "timepoint": 2387, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2387, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; needle,patient,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_878_0", + "timepoint": 878, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 878, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2755_0", + "timepoint": 2755, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2755, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; assistant,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_415_0", + "timepoint": 415, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 415, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,forceps,holding; assistant,forceps,holding; curette,patient,touching; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; assistant,instrument table,looking; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1061_0", + "timepoint": 1061, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1061, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2476_0", + "timepoint": 2476, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2476, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,syringe,injecting; needle,patient,touching; patient,operating table,lyingon; assistant,needle,holding; syringe,needle,touching; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1929_0", + "timepoint": 1929, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1929, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,instrument table,looking; assistant,operating table,closeto; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2375_0", + "timepoint": 2375, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2375, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2514_0", + "timepoint": 2514, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2514, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_1_1_10_0", + "timepoint": 10, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 10, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_1323_0", + "timepoint": 1323, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1323, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1567_0", + "timepoint": 1567, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1567, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1641_0", + "timepoint": 1641, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1641, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,tissue mark,looking; assistant,operating table,closeto; assistant,needle,inserting; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_656_0", + "timepoint": 656, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 656, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_435_0", + "timepoint": 435, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 435, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; assistant,instrument table,looking; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2575_0", + "timepoint": 2575, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2575, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2314_0", + "timepoint": 2314, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2314, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; syringe,needle,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_804_0", + "timepoint": 804, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 804, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_629_0", + "timepoint": 629, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 629, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3227_0", + "timepoint": 3227, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3227, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; assistant,tissue paper,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2057_0", + "timepoint": 2057, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2057, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,syringe,injecting; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3197_0", + "timepoint": 3197, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3197, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; assistant,tissue paper,applying; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_522_0", + "timepoint": 522, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 522, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating room,entering; circulator,assistant,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2542_0", + "timepoint": 2542, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2542, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching; assistant,needle,holding; ultrasound probe,patient,touching; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_618_0", + "timepoint": 618, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 618, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_750_0", + "timepoint": 750, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 750, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1537_0", + "timepoint": 1537, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1537, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_496_0", + "timepoint": 496, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 496, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1182_0", + "timepoint": 1182, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1182, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; assistant,needle,inserting; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_341_0", + "timepoint": 341, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 341, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,looking; circulator,patient,closeto; patient,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_55_0", + "timepoint": 55, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 55, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,scalpel,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2574_0", + "timepoint": 2574, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2574, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_400_0", + "timepoint": 400, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 400, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; circulator,patient,closeto; patient,operating table,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_314_0", + "timepoint": 314, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 314, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound gel,applying; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2582_0", + "timepoint": 2582, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2582, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; assistant,syringe,injecting; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2476_0", + "timepoint": 2476, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2476, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound machine,looking; assistant,instrument table,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_3008_0", + "timepoint": 3008, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3008, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto; tissue paper,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_505_0", + "timepoint": 505, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 505, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3095_0", + "timepoint": 3095, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3095, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,tissue paper,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_778_0", + "timepoint": 778, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 778, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_603_0", + "timepoint": 603, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 603, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_774_0", + "timepoint": 774, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 774, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1221_0", + "timepoint": 1221, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1221, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_3084_0", + "timepoint": 3084, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3084, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_536_0", + "timepoint": 536, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 536, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1186_0", + "timepoint": 1186, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1186, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; circulator,ultrasound machine,manipulating; assistant,antiseptic,holding; circulator,ultrasound machine,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2023_0", + "timepoint": 2023, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2023, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; syringe,needle,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,syringe,injecting; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_777_0", + "timepoint": 777, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 777, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_527_0", + "timepoint": 527, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 527, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1585_0", + "timepoint": 1585, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1585, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_929_0", + "timepoint": 929, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 929, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,body marker,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_549_0", + "timepoint": 549, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 549, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2585_0", + "timepoint": 2585, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2585, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,syringe,injecting; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; syringe,needle,touching; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2030_0", + "timepoint": 2030, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2030, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,ultrasound probe,looking; assistant,syringe,holding; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,needle,inserting; head surgeon,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_3043_0", + "timepoint": 3043, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3043, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; tissue paper,patient,touching; head surgeon,tissue paper,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_392_0", + "timepoint": 392, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 392, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon; circulator,operating table,manipulating; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2060_0", + "timepoint": 2060, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2060, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; assistant,syringe,injecting; assistant,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_121_0", + "timepoint": 121, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 121, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,ultrasound gel,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1419_0", + "timepoint": 1419, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1419, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_545_0", + "timepoint": 545, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 545, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; curette,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; scalpel,patient,touching; head surgeon,curette,holding; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "Ultrasound_5_6_2230_0", + "timepoint": 2230, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2230, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_2150_0", + "timepoint": 2150, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2150, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,patient,disinfection; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; forceps,cotton,touching; assistant,operating table,closeto; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3124_0", + "timepoint": 3124, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3124, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; tissue paper,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2278_0", + "timepoint": 2278, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2278, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,needle,holding; assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1812_0", + "timepoint": 1812, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1812, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1836_0", + "timepoint": 1836, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1836, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; needle,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_510_0", + "timepoint": 510, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 510, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2132_0", + "timepoint": 2132, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2132, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; assistant,needle,holding; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_2_2_437_0", + "timepoint": 437, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 437, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,scalpel,holding; patient,operating table,lyingon; assistant,instrument table,looking; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,cotton,holding; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "MISS_1_1_1968_0", + "timepoint": 1968, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1968, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,instrument table,looking; assistant,antiseptic,holding; head surgeon,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_5_6_2743_0", + "timepoint": 2743, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2743, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,syringe,injecting; needle,patient,touching; syringe,needle,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_350_0", + "timepoint": 350, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 350, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound gel,applying; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2469_0", + "timepoint": 2469, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2469, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_17_0", + "timepoint": 17, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 17, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_1212_0", + "timepoint": 1212, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1212, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_506_0", + "timepoint": 506, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 506, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1720_0", + "timepoint": 1720, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1720, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_1114_0", + "timepoint": 1114, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1114, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; anesthetist,operating table,closeto; assistant,cotton,holding; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_486_0", + "timepoint": 486, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 486, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1300_0", + "timepoint": 1300, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1300, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,patient,manipulating; anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1878_0", + "timepoint": 1878, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1878, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; syringe,needle,touching; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_1813_0", + "timepoint": 1813, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1813, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,antiseptic,holding; assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1123_0", + "timepoint": 1123, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1123, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,body marker,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1728_0", + "timepoint": 1728, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1728, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_528_0", + "timepoint": 528, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 528, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,operating table,manipulating; anesthetist,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1944_0", + "timepoint": 1944, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1944, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,needle,holding; head surgeon,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,needle,holding; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_3_6_393_0", + "timepoint": 393, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 393, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,forceps,holding; head surgeon,forceps,holding; assistant,operating table,closeto; curette,patient,touching; patient,operating table,lyingon; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_682_0", + "timepoint": 682, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 682, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1128_0", + "timepoint": 1128, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1128, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; ultrasound probe,patient,touching; assistant,needle,inserting; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2250_0", + "timepoint": 2250, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2250, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,patient,disinfection; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,forceps,holding; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1496_0", + "timepoint": 1496, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1496, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1747_0", + "timepoint": 1747, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1747, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1582_0", + "timepoint": 1582, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1582, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_575_0", + "timepoint": 575, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 575, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2508_0", + "timepoint": 2508, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2508, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,tissue paper,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_35_0", + "timepoint": 35, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 35, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_94_0", + "timepoint": 94, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 94, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_1_1_84_0", + "timepoint": 84, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 84, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,antiseptic,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_516_0", + "timepoint": 516, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 516, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,controlling; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,operating table,closeto; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1633_0", + "timepoint": 1633, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1633, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2866_0", + "timepoint": 2866, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2866, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; syringe,needle,touching; assistant,syringe,removing; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1838_0", + "timepoint": 1838, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1838, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,needle,inserting; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1428_0", + "timepoint": 1428, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1428, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_3_6_869_0", + "timepoint": 869, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 869, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_988_0", + "timepoint": 988, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 988, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_899_0", + "timepoint": 899, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 899, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,manipulating; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,scissors,holding; assistant,operating table,closeto; scissors,cotton,touching; circulator,ultrasound machine,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "MISS_1_1_950_0", + "timepoint": 950, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 950, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_2240_0", + "timepoint": 2240, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2240, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,needle,holding; assistant,needle,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2071_0", + "timepoint": 2071, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2071, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1854_0", + "timepoint": 1854, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1854, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,syringe,holding; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_684_0", + "timepoint": 684, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 684, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1301_0", + "timepoint": 1301, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1301, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; patient,operating table,lyingon; assistant,antiseptic,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_224_0", + "timepoint": 224, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 224, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_156_0", + "timepoint": 156, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 156, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_257_0", + "timepoint": 257, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 257, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1260_0", + "timepoint": 1260, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1260, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; body marker,patient,touching; assistant,ultrasound probe,looking; assistant,body marker,applying; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_3066_0", + "timepoint": 3066, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3066, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1026_0", + "timepoint": 1026, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1026, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,body marker,holding; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_369_0", + "timepoint": 369, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 369, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; circulator,operating table,closeto; patient,operating table,closeto; circulator,patient,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_531_0", + "timepoint": 531, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 531, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1939_0", + "timepoint": 1939, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1939, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,vertebrae,scanning; syringe,needle,touching; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1811_0", + "timepoint": 1811, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1811, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,needle,holding; ultrasound probe,patient,touching; assistant,syringe,holding; assistant,instrument table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_456_0", + "timepoint": 456, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 456, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,herbal disk,removing; head surgeon,forceps,holding; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; curette,patient,touching; head surgeon,scalpel,holding; assistant,operating table,closeto; head surgeon,curette,holding; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1915_0", + "timepoint": 1915, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1915, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,needle,applying; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_1590_0", + "timepoint": 1590, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1590, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1400_0", + "timepoint": 1400, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1400, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,patient,anaesthetising; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_573_0", + "timepoint": 573, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 573, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_328_0", + "timepoint": 328, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 328, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,manipulating; patient,operating room,entering; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_448_0", + "timepoint": 448, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 448, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2765_0", + "timepoint": 2765, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2765, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; syringe,needle,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2677_0", + "timepoint": 2677, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2677, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2549_0", + "timepoint": 2549, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2549, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,needle,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_3009_0", + "timepoint": 3009, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3009, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; tissue paper,patient,touching; head surgeon,tissue paper,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1633_0", + "timepoint": 1633, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1633, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_815_0", + "timepoint": 815, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 815, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_3_6_490_0", + "timepoint": 490, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 490, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,scalpel,holding; head surgeon,curette,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; curette,patient,touching; head surgeon,herbal disk,removing; head surgeon,microscope eye,looking; scalpel,patient,touching; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1164_0", + "timepoint": 1164, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1164, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1357_0", + "timepoint": 1357, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1357, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_1412_0", + "timepoint": 1412, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1412, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_320_0", + "timepoint": 320, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 320, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; assistant,ultrasound gel,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1425_0", + "timepoint": 1425, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1425, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1426_0", + "timepoint": 1426, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1426, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2198_0", + "timepoint": 2198, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2198, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning; syringe,needle,touching; assistant,syringe,injecting; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2177_0", + "timepoint": 2177, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2177, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; syringe,needle,touching; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1602_0", + "timepoint": 1602, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1602, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_2_2_582_0", + "timepoint": 582, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 582, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,health monitor,looking; head surgeon,microscope controller,controlling; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_35_0", + "timepoint": 35, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 35, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,looking; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_160_0", + "timepoint": 160, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 160, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,cutting; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; forceps,cotton,touching; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,scalpel,holding; scalpel,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_958_0", + "timepoint": 958, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 958, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,body marker,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1835_0", + "timepoint": 1835, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1835, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; assistant,needle,inserting; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_2823_0", + "timepoint": 2823, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2823, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,syringe,removing; assistant,operating table,closeto; syringe,needle,touching; head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1188_0", + "timepoint": 1188, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1188, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,inserting; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2121_0", + "timepoint": 2121, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2121, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; syringe,needle,touching; needle,patient,touching; ultrasound probe,patient,touching; assistant,syringe,injecting; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_687_0", + "timepoint": 687, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 687, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_361_0", + "timepoint": 361, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 361, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,ultrasound gel,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_241_0", + "timepoint": 241, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 241, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_118_0", + "timepoint": 118, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 118, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2087_0", + "timepoint": 2087, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2087, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; needle,patient,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,syringe,injecting; syringe,needle,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1213_0", + "timepoint": 1213, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1213, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2865_0", + "timepoint": 2865, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2865, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,looking; head surgeon,ultrasound machine,looking; patient,operating table,lyingon; assistant,instrument table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2386_0", + "timepoint": 2386, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2386, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; patient,operating table,lyingon; syringe,needle,touching; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1023_0", + "timepoint": 1023, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1023, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1574_0", + "timepoint": 1574, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1574, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_99_0", + "timepoint": 99, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 99, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2320_0", + "timepoint": 2320, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2320, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_252_0", + "timepoint": 252, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 252, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,patient,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1585_0", + "timepoint": 1585, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1585, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,inserting; patient,operating table,lyingon; needle,patient,touching; assistant,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3107_0", + "timepoint": 3107, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3107, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,tissue paper,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_792_0", + "timepoint": 792, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 792, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; assistant,antiseptic,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_503_0", + "timepoint": 503, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 503, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering; anesthetist,operating table,manipulating; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_1270_0", + "timepoint": 1270, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1270, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; assistant,antiseptic,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2691_0", + "timepoint": 2691, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2691, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; syringe,needle,touching; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_686_0", + "timepoint": 686, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 686, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2830_0", + "timepoint": 2830, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2830, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,syringe,removing; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_1173_0", + "timepoint": 1173, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1173, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,manipulating; forceps,cotton,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding; circulator,ultrasound machine,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_405_0", + "timepoint": 405, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 405, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; assistant,instrument table,looking; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "MISS_1_1_2042_0", + "timepoint": 2042, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2042, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; assistant,patient,disinfection; anesthetist,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_182_0", + "timepoint": 182, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 182, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2366_0", + "timepoint": 2366, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2366, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2497_0", + "timepoint": 2497, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2497, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,syringe,injecting; assistant,needle,holding; patient,operating table,lyingon; head surgeon,needle,holding; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_1_1_2142_0", + "timepoint": 2142, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2142, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; cotton,patient,touching; assistant,forceps,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,patient,disinfection; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3146_0", + "timepoint": 3146, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3146, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,applying; tissue paper,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_389_0", + "timepoint": 389, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 389, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,patient,aspirating; cotton,patient,touching; forceps,cotton,touching; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,scalpel,holding; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_310_0", + "timepoint": 310, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 310, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,health monitor,looking; patient,operating table,lyingon; assistant,patient,aspirating; forceps,cotton,touching; assistant,operating table,closeto; head surgeon,scalpel,holding; head surgeon,operating table,closeto; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_294_0", + "timepoint": 294, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 294, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,patient,looking " + } + ] + }, + { + "id": "MISS_1_1_240_0", + "timepoint": 240, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 240, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2207_0", + "timepoint": 2207, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2207, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,needle,holding; head surgeon,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1577_0", + "timepoint": 1577, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1577, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1655_0", + "timepoint": 1655, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1655, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,tissue mark,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,inserting; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_2_2_1055_0", + "timepoint": 1055, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1055, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1678_0", + "timepoint": 1678, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1678, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_84_0", + "timepoint": 84, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 84, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,scalpel,holding; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_2769_0", + "timepoint": 2769, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2769, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; syringe,needle,touching; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,needle,holding; assistant,syringe,injecting " + } + ] + }, + { + "id": "Ultrasound_4_5_96_0", + "timepoint": 96, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 96, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_476_0", + "timepoint": 476, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 476, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2925_0", + "timepoint": 2925, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2925, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,syringe,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,needle,holding; needle,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_1103_0", + "timepoint": 1103, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1103, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; circulator,ultrasound machine,closeto; assistant,forceps,holding; forceps,cotton,touching; assistant,antiseptic,holding; assistant,operating table,closeto; circulator,ultrasound machine,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2529_0", + "timepoint": 2529, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2529, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2406_0", + "timepoint": 2406, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2406, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,removing; head surgeon,ultrasound machine,looking; needle,patient,touching; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_449_0", + "timepoint": 449, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 449, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1318_0", + "timepoint": 1318, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1318, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,inserting; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1359_0", + "timepoint": 1359, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1359, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,removing; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_149_0", + "timepoint": 149, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 149, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,applying; assistant,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_164_0", + "timepoint": 164, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 164, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2078_0", + "timepoint": 2078, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2078, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; forceps,cotton,touching; assistant,patient,disinfection; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1529_0", + "timepoint": 1529, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1529, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_185_0", + "timepoint": 185, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 185, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2464_0", + "timepoint": 2464, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2464, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,patient,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2067_0", + "timepoint": 2067, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2067, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; needle,patient,touching; syringe,needle,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1069_0", + "timepoint": 1069, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1069, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,body marker,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2005_0", + "timepoint": 2005, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2005, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; syringe,needle,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1511_0", + "timepoint": 1511, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1511, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,ultrasound probe,looking; assistant,needle,inserting; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_943_0", + "timepoint": 943, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 943, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_1922_0", + "timepoint": 1922, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1922, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,patient,disinfection; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,applying " + } + ] + }, + { + "id": "Ultrasound_3_4_1769_0", + "timepoint": 1769, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1769, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,tissue mark,looking; patient,operating table,lyingon; assistant,needle,inserting; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1328_0", + "timepoint": 1328, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1328, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,patient,manipulating; anesthetist,anesthesia equipment,holding; assistant,instrument table,looking; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_295_0", + "timepoint": 295, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 295, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1859_0", + "timepoint": 1859, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1859, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,patient,disinfection; head surgeon,operating table,closeto; assistant,antiseptic,applying; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_595_0", + "timepoint": 595, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 595, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_2578_0", + "timepoint": 2578, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2578, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; needle,patient,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; syringe,needle,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_769_0", + "timepoint": 769, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 769, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,instrument table,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2959_0", + "timepoint": 2959, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2959, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; head surgeon,tissue paper,applying; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1140_0", + "timepoint": 1140, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1140, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating room,entering; head surgeon,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_108_0", + "timepoint": 108, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 108, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1834_0", + "timepoint": 1834, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1834, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; head surgeon,vertebrae,scanning; assistant,needle,inserting; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1080_0", + "timepoint": 1080, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1080, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,manipulating; anesthetist,microscope,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_110_0", + "timepoint": 110, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 110, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; head surgeon,patient,manipulating; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1514_0", + "timepoint": 1514, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1514, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,inserting; assistant,ultrasound probe,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_765_0", + "timepoint": 765, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 765, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1172_0", + "timepoint": 1172, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1172, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,body marker,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_257_0", + "timepoint": 257, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 257, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; circulator,patient,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2852_0", + "timepoint": 2852, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2852, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,removing; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,needle,holding; patient,operating table,lyingon; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_394_0", + "timepoint": 394, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 394, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2112_0", + "timepoint": 2112, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2112, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,syringe,injecting; head surgeon,operating table,closeto; syringe,needle,touching; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_1375_0", + "timepoint": 1375, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1375, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; anesthetist,operating table,closeto; assistant,instrument table,looking; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; circulator,operating table,looking; patient,operating table,lyingon; anesthetist,patient,anaesthetising " + } + ] + }, + { + "id": "Ultrasound_4_5_2331_0", + "timepoint": 2331, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2331, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_3_6_307_0", + "timepoint": 307, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 307, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; anesthetist,health monitor,looking; head surgeon,curette,holding; patient,operating table,lyingon; curette,patient,touching; anesthetist,anesthesia equipment,controlling; head surgeon,forceps,holding; forceps,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_769_0", + "timepoint": 769, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 769, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,instruments,disinfection; patient,operating table,lyingon; anesthetist,health monitor,looking; assistant,anesthetist,applying " + } + ] + }, + { + "id": "Ultrasound_3_4_854_0", + "timepoint": 854, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 854, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; body marker,patient,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,body marker,applying; patient,operating table,lyingon; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_918_0", + "timepoint": 918, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 918, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,antiseptic,holding; forceps,cotton,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,patient,disinfection; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1057_0", + "timepoint": 1057, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1057, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2058_0", + "timepoint": 2058, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2058, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,syringe,injecting; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,needle,holding; head surgeon,vertebrae,scanning; syringe,needle,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_704_0", + "timepoint": 704, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 704, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1519_0", + "timepoint": 1519, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1519, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope controller,manipulating; assistant,operating table,closeto; anesthetist,patient,anaesthetising; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1478_0", + "timepoint": 1478, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1478, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,microscope controller,manipulating; anesthetist,patient,anaesthetising; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_582_0", + "timepoint": 582, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 582, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_464_0", + "timepoint": 464, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 464, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,forceps,holding; curette,patient,touching; head surgeon,herbal disk,removing; patient,operating table,lyingon; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; head surgeon,curette,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1103_0", + "timepoint": 1103, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1103, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,body marker,holding; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1886_0", + "timepoint": 1886, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1886, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound screen,looking; assistant,syringe,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_261_0", + "timepoint": 261, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 261, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_112_0", + "timepoint": 112, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 112, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,syringe,looking; circulator,instrument table,closeto; circulator,instruments,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2208_0", + "timepoint": 2208, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2208, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2324_0", + "timepoint": 2324, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2324, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_358_0", + "timepoint": 358, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 358, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; cotton,patient,touching; anesthetist,operating table,closeto; anesthetist,health monitor,looking; forceps,cotton,touching; head surgeon,scalpel,holding; assistant,patient,aspirating; patient,operating table,lyingon; assistant,forceps,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3205_0", + "timepoint": 3205, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3205, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,applying; head surgeon,operating table,closeto; tissue paper,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_813_0", + "timepoint": 813, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 813, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; circulator,microscope,manipulating; circulator,microscope,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2296_0", + "timepoint": 2296, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2296, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; head surgeon,vertebrae,scanning; patient,operating table,lyingon; needle,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_1015_0", + "timepoint": 1015, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1015, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,manipulating; patient,operating table,lyingon; anesthetist,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_1143_0", + "timepoint": 1143, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1143, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,body marker,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_284_0", + "timepoint": 284, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 284, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,looking; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_1765_0", + "timepoint": 1765, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1765, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,syringe,holding; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2012_0", + "timepoint": 2012, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2012, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,syringe,holding; assistant,ultrasound probe,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,needle,inserting; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1701_0", + "timepoint": 1701, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1701, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_277_0", + "timepoint": 277, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 277, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,manipulating; patient,operating table,lyingon; patient,operating room,entering; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_2466_0", + "timepoint": 2466, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2466, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,needle,holding; needle,patient,touching; assistant,syringe,holding; assistant,instrument table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_183_0", + "timepoint": 183, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 183, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,curette,holding; anesthetist,operating table,closeto; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; curette,patient,touching; assistant,operating table,closeto; head surgeon,forceps,holding " + } + ] + }, + { + "id": "MISS_1_1_706_0", + "timepoint": 706, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 706, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_177_0", + "timepoint": 177, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 177, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,forceps,holding; patient,operating table,lyingon; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; curette,patient,touching; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1250_0", + "timepoint": 1250, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1250, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_726_0", + "timepoint": 726, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 726, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,patient,manipulating; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2251_0", + "timepoint": 2251, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2251, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,syringe,injecting; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; syringe,needle,touching; assistant,operating table,closeto; assistant,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_1162_0", + "timepoint": 1162, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1162, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; circulator,ultrasound machine,manipulating; patient,operating table,lyingon; assistant,antiseptic,holding; circulator,ultrasound machine,closeto; forceps,cotton,touching; assistant,forceps,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1132_0", + "timepoint": 1132, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1132, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,closeto; patient,operating table,lyingon; assistant,forceps,holding; assistant,operating table,closeto; assistant,antiseptic,holding; circulator,ultrasound machine,manipulating; forceps,cotton,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2881_0", + "timepoint": 2881, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2881, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; assistant,syringe,removing; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_2_2_86_0", + "timepoint": 86, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 86, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,scalpel,holding; anesthetist,operating table,closeto; head surgeon,patient,manipulating; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3076_0", + "timepoint": 3076, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3076, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; head surgeon,operating table,closeto; assistant,tissue paper,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_446_0", + "timepoint": 446, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 446, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,instrument table,looking; assistant,forceps,holding; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,scalpel,holding; assistant,cotton,holding; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "MISS_3_6_735_0", + "timepoint": 735, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 735, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; head surgeon,microscope eye,looking; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,anesthetist,applying; anesthetist,anesthesia equipment,controlling; assistant,instruments,disinfection; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_339_0", + "timepoint": 339, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 339, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2847_0", + "timepoint": 2847, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2847, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; needle,patient,touching; assistant,syringe,removing; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_462_0", + "timepoint": 462, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 462, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_28_0", + "timepoint": 28, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 28, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1590_0", + "timepoint": 1590, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1590, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2817_0", + "timepoint": 2817, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2817, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,holding; syringe,needle,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,syringe,removing " + } + ] + }, + { + "id": "MISS_2_2_120_0", + "timepoint": 120, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 120, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,scalpel,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; head surgeon,patient,cutting " + } + ] + }, + { + "id": "MISS_3_6_950_0", + "timepoint": 950, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 950, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; cotton,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,patient,aspirating; anesthetist,anesthesia equipment,controlling; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_1142_0", + "timepoint": 1142, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1142, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching; assistant,antiseptic,holding; circulator,ultrasound machine,closeto; assistant,operating table,closeto; circulator,ultrasound machine,manipulating; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_1_1_924_0", + "timepoint": 924, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 924, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,manipulating; anesthetist,microscope,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1064_0", + "timepoint": 1064, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1064, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1196_0", + "timepoint": 1196, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1196, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3096_0", + "timepoint": 3096, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3096, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,tissue paper,holding; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1216_0", + "timepoint": 1216, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1216, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2871_0", + "timepoint": 2871, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2871, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound machine,looking; assistant,needle,holding; patient,operating table,lyingon; assistant,instrument table,looking; assistant,instrument table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1570_0", + "timepoint": 1570, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1570, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2095_0", + "timepoint": 2095, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2095, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; assistant,syringe,injecting; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1603_0", + "timepoint": 1603, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1603, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_845_0", + "timepoint": 845, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 845, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1618_0", + "timepoint": 1618, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1618, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1936_0", + "timepoint": 1936, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1936, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,antiseptic,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1142_0", + "timepoint": 1142, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1142, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2920_0", + "timepoint": 2920, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2920, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,holding; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,syringe,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_509_0", + "timepoint": 509, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 509, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; curette,patient,touching; head surgeon,scalpel,holding; anesthetist,anesthesia equipment,controlling; head surgeon,curette,holding; head surgeon,microscope eye,looking; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,herbal disk,removing; scalpel,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2682_0", + "timepoint": 2682, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2682, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,needle,holding; syringe,needle,touching; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_2672_0", + "timepoint": 2672, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2672, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; syringe,needle,touching; head surgeon,ultrasound screen,looking; assistant,syringe,injecting; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2917_0", + "timepoint": 2917, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2917, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_49_0", + "timepoint": 49, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 49, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking; circulator,scalpel,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_1243_0", + "timepoint": 1243, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1243, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; patient,operating table,lyingon; assistant,antiseptic,holding " + } + ] + }, + { + "id": "MISS_1_1_677_0", + "timepoint": 677, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 677, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2562_0", + "timepoint": 2562, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2562, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_328_0", + "timepoint": 328, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 328, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; assistant,ultrasound gel,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1847_0", + "timepoint": 1847, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1847, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; assistant,needle,holding; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1787_0", + "timepoint": 1787, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1787, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,syringe,holding; assistant,instrument table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; head surgeon,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2527_0", + "timepoint": 2527, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2527, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; head surgeon,operating table,closeto; assistant,patient,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_84_0", + "timepoint": 84, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 84, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,operating table,closeto; patient,operating table,lyingon; curette,patient,touching; head surgeon,operating table,closeto; head surgeon,herbal disk,removing; assistant,operating table,closeto; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_340_0", + "timepoint": 340, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 340, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_4_5_2073_0", + "timepoint": 2073, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2073, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; needle,patient,touching; syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1289_0", + "timepoint": 1289, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1289, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_409_0", + "timepoint": 409, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 409, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1795_0", + "timepoint": 1795, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1795, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,inserting; needle,patient,touching; assistant,operating table,closeto; assistant,tissue mark,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_1971_0", + "timepoint": 1971, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1971, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,syringe,injecting; assistant,operating table,closeto; needle,patient,touching; head surgeon,needle,holding; syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_2467_0", + "timepoint": 2467, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2467, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,needle,holding; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1592_0", + "timepoint": 1592, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1592, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; assistant,needle,inserting; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_286_0", + "timepoint": 286, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 286, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound gel,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2202_0", + "timepoint": 2202, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2202, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,syringe,injecting; head surgeon,operating table,closeto; assistant,needle,holding; head surgeon,needle,holding; ultrasound probe,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_967_0", + "timepoint": 967, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 967, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2547_0", + "timepoint": 2547, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2547, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; syringe,needle,touching; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_74_0", + "timepoint": 74, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 74, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,patient,manipulating; assistant,scalpel,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_86_0", + "timepoint": 86, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 86, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,curette,looking; circulator,instrument table,closeto; circulator,instruments,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_235_0", + "timepoint": 235, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 235, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; head surgeon,ultrasound probe,controlling; assistant,instrument table,looking; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1998_0", + "timepoint": 1998, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1998, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,syringe,injecting; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,needle,holding; syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1911_0", + "timepoint": 1911, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1911, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,applying; assistant,patient,disinfection; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_388_0", + "timepoint": 388, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 388, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound gel,holding; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1369_0", + "timepoint": 1369, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1369, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; assistant,instrument table,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; anesthetist,operating table,closeto; circulator,operating table,looking; anesthetist,patient,anaesthetising " + } + ] + }, + { + "id": "MISS_1_1_1138_0", + "timepoint": 1138, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1138, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating room,entering; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating room,entering " + } + ] + }, + { + "id": "MISS_3_6_273_0", + "timepoint": 273, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 273, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; assistant,operating table,closeto; head surgeon,forceps,holding; forceps,patient,touching; head surgeon,microscope eye,looking; curette,patient,touching; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2899_0", + "timepoint": 2899, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2899, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound machine,looking; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_107_0", + "timepoint": 107, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 107, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; head surgeon,patient,manipulating; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_818_0", + "timepoint": 818, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 818, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,body marker,holding; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_653_0", + "timepoint": 653, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 653, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_2032_0", + "timepoint": 2032, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2032, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; assistant,syringe,holding; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_799_0", + "timepoint": 799, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 799, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,instrument table,closeto; assistant,forceps,holding; forceps,cotton,touching; assistant,antiseptic,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_97_0", + "timepoint": 97, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 97, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_353_0", + "timepoint": 353, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 353, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1463_0", + "timepoint": 1463, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1463, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_816_0", + "timepoint": 816, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 816, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,body marker,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1436_0", + "timepoint": 1436, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1436, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2287_0", + "timepoint": 2287, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2287, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_341_0", + "timepoint": 341, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 341, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,ultrasound gel,applying; head surgeon,ultrasound probe,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_442_0", + "timepoint": 442, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 442, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,closeto; assistant,operating room,entering " + } + ] + }, + { + "id": "MISS_3_6_886_0", + "timepoint": 886, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 886, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; forceps,cotton,touching; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1389_0", + "timepoint": 1389, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1389, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_411_0", + "timepoint": 411, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 411, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto; assistant,instrument table,looking; patient,operating table,lyingon; forceps,cotton,touching; head surgeon,scalpel,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3221_0", + "timepoint": 3221, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3221, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,tissue paper,applying; tissue paper,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_756_0", + "timepoint": 756, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 756, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; assistant,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1552_0", + "timepoint": 1552, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1552, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_783_0", + "timepoint": 783, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 783, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2419_0", + "timepoint": 2419, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2419, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,instrument table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,needle,holding; assistant,instrument table,looking; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2433_0", + "timepoint": 2433, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2433, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; syringe,needle,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; needle,patient,touching; assistant,needle,holding; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_2195_0", + "timepoint": 2195, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2195, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; needle,patient,touching; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,holding; assistant,operating table,closeto; syringe,needle,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2683_0", + "timepoint": 2683, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2683, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,needle,holding; assistant,syringe,injecting; assistant,needle,holding; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_3_6_853_0", + "timepoint": 853, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 853, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; assistant,forceps,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_723_0", + "timepoint": 723, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 723, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_195_0", + "timepoint": 195, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 195, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,patient,manipulating; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_240_0", + "timepoint": 240, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 240, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto " + } + ] + }, + { + "id": "MISS_3_6_611_0", + "timepoint": 611, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 611, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; curette,patient,touching; patient,operating table,lyingon; head surgeon,curette,holding; assistant,operating table,closeto; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; scalpel,patient,touching; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "MISS_1_1_1665_0", + "timepoint": 1665, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1665, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_714_0", + "timepoint": 714, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 714, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1511_0", + "timepoint": 1511, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1511, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,needle,holding; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2712_0", + "timepoint": 2712, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2712, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,needle,holding; assistant,syringe,injecting; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; syringe,needle,touching; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1299_0", + "timepoint": 1299, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1299, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,needle,inserting; patient,operating table,lyingon; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_944_0", + "timepoint": 944, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 944, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_482_0", + "timepoint": 482, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 482, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon; anesthetist,operating table,manipulating; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_2878_0", + "timepoint": 2878, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2878, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; head surgeon,operating table,closeto; assistant,needle,holding; patient,operating table,lyingon; head surgeon,ultrasound machine,looking; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_2_2_187_0", + "timepoint": 187, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 187, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " scalpel,patient,touching; head surgeon,scalpel,holding; assistant,forceps,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,cutting; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_276_0", + "timepoint": 276, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 276, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_406_0", + "timepoint": 406, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 406, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,touching; circulator,operating table,closeto; circulator,patient,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1285_0", + "timepoint": 1285, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1285, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; body marker,patient,touching; assistant,body marker,applying " + } + ] + }, + { + "id": "Ultrasound_4_5_177_0", + "timepoint": 177, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 177, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_30_0", + "timepoint": 30, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 30, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_155_0", + "timepoint": 155, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 155, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_356_0", + "timepoint": 356, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 356, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound gel,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_3181_0", + "timepoint": 3181, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3181, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,tissue paper,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; tissue paper,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_964_0", + "timepoint": 964, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 964, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,forceps,holding; cotton,patient,touching; assistant,patient,disinfection; assistant,antiseptic,holding " + } + ] + }, + { + "id": "MISS_3_6_824_0", + "timepoint": 824, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 824, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; assistant,instrument table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1306_0", + "timepoint": 1306, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1306, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,holding; patient,operating table,lyingon; anesthetist,patient,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1530_0", + "timepoint": 1530, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1530, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_280_0", + "timepoint": 280, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 280, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1319_0", + "timepoint": 1319, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1319, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2438_0", + "timepoint": 2438, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2438, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound machine,looking; assistant,needle,removing; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2936_0", + "timepoint": 2936, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2936, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,tissue paper,applying " + } + ] + }, + { + "id": "Ultrasound_3_4_2568_0", + "timepoint": 2568, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2568, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,syringe,injecting; head surgeon,needle,holding; needle,patient,touching; assistant,operating table,closeto; assistant,needle,holding; syringe,needle,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_831_0", + "timepoint": 831, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 831, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; assistant,body marker,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2081_0", + "timepoint": 2081, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2081, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; assistant,patient,disinfection; head surgeon,operating table,closeto; forceps,cotton,touching; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_829_0", + "timepoint": 829, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 829, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,body marker,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2060_0", + "timepoint": 2060, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2060, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; head surgeon,operating table,closeto; assistant,operating table,closeto; syringe,needle,touching; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1815_0", + "timepoint": 1815, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1815, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,antiseptic,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_430_0", + "timepoint": 430, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 430, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2485_0", + "timepoint": 2485, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2485, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; syringe,needle,touching; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_1_1_2045_0", + "timepoint": 2045, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2045, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; assistant,patient,disinfection; patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1827_0", + "timepoint": 1827, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1827, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; needle,patient,touching; assistant,syringe,injecting; head surgeon,needle,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; syringe,needle,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_29_0", + "timepoint": 29, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 29, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,patient,manipulating; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2579_0", + "timepoint": 2579, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2579, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,tissue paper,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_272_0", + "timepoint": 272, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 272, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound gel,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2224_0", + "timepoint": 2224, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2224, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; syringe,needle,touching; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,needle,holding; assistant,syringe,injecting; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1554_0", + "timepoint": 1554, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1554, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2814_0", + "timepoint": 2814, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2814, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound machine,looking; needle,patient,touching; head surgeon,operating table,closeto; assistant,needle,removing; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2473_0", + "timepoint": 2473, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2473, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound machine,looking; assistant,needle,holding; assistant,instrument table,looking; assistant,instrument table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2040_0", + "timepoint": 2040, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2040, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; needle,patient,touching; assistant,operating table,closeto; assistant,syringe,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_838_0", + "timepoint": 838, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 838, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1283_0", + "timepoint": 1283, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1283, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,holding; anesthetist,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2124_0", + "timepoint": 2124, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2124, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,needle,holding; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; syringe,needle,touching; assistant,ultrasound screen,looking; assistant,syringe,injecting; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_556_0", + "timepoint": 556, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 556, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1679_0", + "timepoint": 1679, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1679, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,instrument table,looking; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_412_0", + "timepoint": 412, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 412, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; anesthetist,operating table,manipulating; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_2913_0", + "timepoint": 2913, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2913, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_664_0", + "timepoint": 664, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 664, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_343_0", + "timepoint": 343, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 343, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound gel,applying; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_1_1_1964_0", + "timepoint": 1964, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1964, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,antiseptic,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_3_4_2510_0", + "timepoint": 2510, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2510, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; assistant,operating table,closeto; assistant,needle,holding; head surgeon,needle,holding; needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_292_0", + "timepoint": 292, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 292, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,looking; patient,operating room,entering " + } + ] + }, + { + "id": "MISS_2_2_526_0", + "timepoint": 526, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 526, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,controlling; assistant,operating table,closeto; anesthetist,health monitor,looking; anesthetist,operating table,closeto; circulator,instrument table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_740_0", + "timepoint": 740, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 740, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_3_6_784_0", + "timepoint": 784, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 784, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_378_0", + "timepoint": 378, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 378, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_683_0", + "timepoint": 683, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 683, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_2455_0", + "timepoint": 2455, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2455, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; syringe,needle,touching; head surgeon,needle,holding; assistant,syringe,injecting; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound probe,controlling; assistant,needle,holding; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_302_0", + "timepoint": 302, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 302, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,ultrasound gel,applying; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2532_0", + "timepoint": 2532, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2532, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; anesthetist,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_396_0", + "timepoint": 396, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 396, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,forceps,holding; head surgeon,herbal disk,removing; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; curette,patient,touching; head surgeon,microscope eye,looking; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1619_0", + "timepoint": 1619, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1619, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; assistant,operating table,closeto; assistant,needle,inserting; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2423_0", + "timepoint": 2423, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2423, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,needle,holding; syringe,needle,touching; patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; head surgeon,vertebrae,scanning; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1198_0", + "timepoint": 1198, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1198, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,forceps,holding; circulator,ultrasound machine,manipulating; assistant,antiseptic,holding; forceps,cotton,touching; circulator,ultrasound machine,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_954_0", + "timepoint": 954, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 954, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; head surgeon,microscope eye,looking; forceps,cotton,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; cotton,patient,touching; patient,operating table,lyingon; assistant,forceps,holding; anesthetist,operating table,closeto; assistant,patient,aspirating " + } + ] + }, + { + "id": "MISS_3_6_459_0", + "timepoint": 459, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 459, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; head surgeon,forceps,holding; head surgeon,curette,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "Ultrasound_5_6_1080_0", + "timepoint": 1080, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1080, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_518_0", + "timepoint": 518, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 518, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; anesthetist,operating table,manipulating; circulator,operating table,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2449_0", + "timepoint": 2449, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2449, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_733_0", + "timepoint": 733, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 733, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,microscope screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_1545_0", + "timepoint": 1545, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1545, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_2139_0", + "timepoint": 2139, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2139, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; needle,patient,touching; ultrasound probe,patient,touching; assistant,syringe,injecting; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; syringe,needle,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1274_0", + "timepoint": 1274, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1274, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; body marker,patient,touching; assistant,body marker,applying; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "MISS_1_1_1790_0", + "timepoint": 1790, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1790, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,antiseptic,holding; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2901_0", + "timepoint": 2901, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2901, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "MISS_3_6_1195_0", + "timepoint": 1195, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1195, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1242_0", + "timepoint": 1242, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1242, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; assistant,operating table,closeto; assistant,body marker,applying; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_3000_0", + "timepoint": 3000, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 3000, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,tissue paper,applying; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2500_0", + "timepoint": 2500, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2500, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,manipulating; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,patient,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_789_0", + "timepoint": 789, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 789, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,microscope screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,health monitor,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1357_0", + "timepoint": 1357, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1357, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1620_0", + "timepoint": 1620, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1620, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_194_0", + "timepoint": 194, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 194, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,manipulating; head surgeon,ultrasound probe,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_294_0", + "timepoint": 294, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 294, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon; assistant,ultrasound gel,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_2494_0", + "timepoint": 2494, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2494, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,syringe,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_879_0", + "timepoint": 879, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 879, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1182_0", + "timepoint": 1182, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1182, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_585_0", + "timepoint": 585, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 585, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering; circulator,operating table,manipulating " + } + ] + }, + { + "id": "MISS_1_1_1009_0", + "timepoint": 1009, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1009, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,microscope,manipulating; anesthetist,operating table,closeto; circulator,microscope,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1277_0", + "timepoint": 1277, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1277, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; needle,patient,touching; patient,operating table,lyingon; assistant,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,inserting; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1611_0", + "timepoint": 1611, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1611, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_625_0", + "timepoint": 625, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 625, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_117_0", + "timepoint": 117, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 117, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,scalpel,holding; patient,operating table,lyingon; head surgeon,patient,cutting; assistant,forceps,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_1049_0", + "timepoint": 1049, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1049, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,antiseptic,holding; patient,operating table,lyingon; assistant,operating table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1195_0", + "timepoint": 1195, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1195, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; assistant,needle,inserting " + } + ] + }, + { + "id": "MISS_2_2_195_0", + "timepoint": 195, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 195, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; scalpel,patient,touching; anesthetist,health monitor,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,patient,cutting; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_2655_0", + "timepoint": 2655, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2655, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; ultrasound probe,patient,touching; assistant,needle,holding; patient,operating table,lyingon; head surgeon,needle,holding; assistant,operating table,closeto; syringe,needle,touching; head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; assistant,syringe,injecting; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_3_4_1481_0", + "timepoint": 1481, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1481, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,instrument table,closeto; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_300_0", + "timepoint": 300, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 300, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; assistant,operating table,closeto; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; forceps,patient,touching; head surgeon,microscope eye,looking; curette,patient,touching; anesthetist,health monitor,looking; head surgeon,forceps,holding; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2392_0", + "timepoint": 2392, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2392, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1795_0", + "timepoint": 1795, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1795, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching; head surgeon,needle,holding; assistant,syringe,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1657_0", + "timepoint": 1657, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1657, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; assistant,instrument table,looking; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_678_0", + "timepoint": 678, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 678, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_963_0", + "timepoint": 963, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 963, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_722_0", + "timepoint": 722, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 722, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1300_0", + "timepoint": 1300, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1300, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " body marker,patient,touching; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; assistant,body marker,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1054_0", + "timepoint": 1054, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1054, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,microscope,manipulating; circulator,microscope,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_606_0", + "timepoint": 606, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 606, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1197_0", + "timepoint": 1197, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1197, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,needle,inserting; needle,patient,touching; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2096_0", + "timepoint": 2096, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2096, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,forceps,holding; assistant,patient,disinfection; forceps,cotton,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_174_0", + "timepoint": 174, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 174, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; forceps,cotton,touching; assistant,operating table,closeto; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; scalpel,patient,touching; head surgeon,patient,cutting; head surgeon,scalpel,holding; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2422_0", + "timepoint": 2422, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2422, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,needle,removing; patient,operating table,lyingon; needle,patient,touching; head surgeon,ultrasound machine,looking; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1500_0", + "timepoint": 1500, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1500, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_518_0", + "timepoint": 518, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 518, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1723_0", + "timepoint": 1723, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1723, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue mark,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; assistant,needle,inserting; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2538_0", + "timepoint": 2538, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2538, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,tissue paper,applying; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1251_0", + "timepoint": 1251, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1251, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,ultrasound probe,looking; assistant,body marker,applying; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2449_0", + "timepoint": 2449, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2449, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,patient,manipulating; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_1179_0", + "timepoint": 1179, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1179, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,needle,inserting; assistant,ultrasound probe,looking; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2187_0", + "timepoint": 2187, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2187, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,needle,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; assistant,operating table,closeto; assistant,syringe,injecting " + } + ] + }, + { + "id": "MISS_1_1_1611_0", + "timepoint": 1611, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1611, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,holding; head surgeon,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1426_0", + "timepoint": 1426, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1426, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,patient,anaesthetising; anesthetist,operating table,closeto; head surgeon,microscope controller,manipulating; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_68_0", + "timepoint": 68, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 68, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,patient,manipulating; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_1170_0", + "timepoint": 1170, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1170, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking " + } + ] + }, + { + "id": "MISS_1_1_2173_0", + "timepoint": 2173, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2173, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; forceps,cotton,touching; cotton,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; assistant,forceps,holding; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_3_4_2740_0", + "timepoint": 2740, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2740, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,needle,holding; needle,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2415_0", + "timepoint": 2415, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2415, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_53_0", + "timepoint": 53, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 53, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,scalpel,looking; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_2_2_636_0", + "timepoint": 636, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 636, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; assistant,microscope screen,looking; assistant,operating table,closeto; head surgeon,microscope controller,controlling; head surgeon,operating table,closeto; head surgeon,microscope screen,looking " + } + ] + }, + { + "id": "MISS_1_1_444_0", + "timepoint": 444, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 444, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering; circulator,operating table,manipulating; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2297_0", + "timepoint": 2297, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2297, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; needle,patient,touching; patient,operating table,lyingon; assistant,syringe,holding; head surgeon,operating table,closeto; head surgeon,needle,holding; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1183_0", + "timepoint": 1183, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1183, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,body marker,holding; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_892_0", + "timepoint": 892, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 892, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,ultrasound screen,looking; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2199_0", + "timepoint": 2199, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2199, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,syringe,injecting; head surgeon,operating table,closeto; assistant,operating table,closeto; needle,patient,touching; patient,operating table,lyingon; syringe,needle,touching; head surgeon,ultrasound screen,looking; head surgeon,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_452_0", + "timepoint": 452, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 452, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,herbal disk,removing; assistant,scalpel,holding; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; curette,patient,touching; assistant,operating table,closeto; head surgeon,forceps,holding; head surgeon,microscope eye,looking; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1073_0", + "timepoint": 1073, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1073, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; circulator,ultrasound machine,looking; assistant,forceps,holding; assistant,antiseptic,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_681_0", + "timepoint": 681, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 681, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2099_0", + "timepoint": 2099, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2099, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; cotton,patient,touching; assistant,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; assistant,patient,disinfection; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2758_0", + "timepoint": 2758, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2758, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,syringe,injecting; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound screen,looking; syringe,needle,touching; patient,operating table,lyingon; assistant,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_815_0", + "timepoint": 815, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 815, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_774_0", + "timepoint": 774, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 774, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_3215_0", + "timepoint": 3215, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3215, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,applying; assistant,operating table,closeto; tissue paper,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_2012_0", + "timepoint": 2012, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2012, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; patient,operating table,lyingon; syringe,needle,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,operating table,closeto; needle,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_2712_0", + "timepoint": 2712, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2712, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,bin,closeto; head surgeon,operating table,closeto; assistant,bin,looking; patient,operating table,lyingon; assistant,tissue paper,dropping " + } + ] + }, + { + "id": "MISS_2_2_146_0", + "timepoint": 146, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 146, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; head surgeon,scalpel,holding; anesthetist,operating table,closeto; scalpel,patient,touching; assistant,instrument table,closeto; head surgeon,patient,cutting; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_718_0", + "timepoint": 718, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 718, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; patient,operating table,lyingon; circulator,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2191_0", + "timepoint": 2191, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2191, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,patient,disinfection; assistant,operating table,closeto; cotton,patient,touching; head surgeon,operating table,closeto; anesthetist,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_959_0", + "timepoint": 959, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 959, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1939_0", + "timepoint": 1939, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1939, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto; assistant,antiseptic,applying; assistant,patient,disinfection; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_3023_0", + "timepoint": 3023, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3023, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,instrument table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,tissue paper,holding; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_449_0", + "timepoint": 449, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 449, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,looking; anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,cotton,holding; anesthetist,anesthesia equipment,controlling; anesthetist,health monitor,looking; head surgeon,scalpel,holding; assistant,forceps,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_10_0", + "timepoint": 10, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 10, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering " + } + ] + }, + { + "id": "MISS_1_1_26_0", + "timepoint": 26, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 26, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating room,entering " + } + ] + }, + { + "id": "MISS_1_1_2370_0", + "timepoint": 2370, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2370, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_931_0", + "timepoint": 931, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 931, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_2082_0", + "timepoint": 2082, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2082, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,disinfection; assistant,operating table,closeto; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1168_0", + "timepoint": 1168, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1168, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; circulator,ultrasound machine,manipulating; forceps,cotton,touching; circulator,ultrasound machine,closeto; assistant,antiseptic,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_137_0", + "timepoint": 137, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 137, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound machine,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1804_0", + "timepoint": 1804, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1804, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,syringe,holding; head surgeon,operating table,closeto; assistant,instrument table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_720_0", + "timepoint": 720, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 720, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1757_0", + "timepoint": 1757, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1757, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; needle,patient,touching; assistant,syringe,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2554_0", + "timepoint": 2554, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2554, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,patient,manipulating; anesthetist,operating table,closeto; assistant,operating table,closeto; assistant,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_65_0", + "timepoint": 65, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 65, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2199_0", + "timepoint": 2199, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2199, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,needle,holding; syringe,needle,touching; head surgeon,needle,holding; needle,patient,touching; head surgeon,vertebrae,scanning; assistant,syringe,injecting; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_57_0", + "timepoint": 57, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 57, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,curette,holding; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_2_2_692_0", + "timepoint": 692, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 692, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope controller,controlling; assistant,operating table,closeto; head surgeon,microscope screen,looking; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,microscope screen,looking " + } + ] + }, + { + "id": "MISS_2_2_83_0", + "timepoint": 83, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 83, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,scalpel,holding; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "MISS_1_1_815_0", + "timepoint": 815, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 815, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; circulator,microscope,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2088_0", + "timepoint": 2088, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2088, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; assistant,syringe,injecting; head surgeon,vertebrae,scanning; assistant,operating table,closeto; syringe,needle,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1155_0", + "timepoint": 1155, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1155, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,ultrasound machine,closeto; assistant,antiseptic,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; circulator,ultrasound machine,manipulating " + } + ] + }, + { + "id": "Ultrasound_2_2_651_0", + "timepoint": 651, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 651, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,patient,manipulating; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_4_5_2682_0", + "timepoint": 2682, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2682, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,tissue paper,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_685_0", + "timepoint": 685, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 685, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1003_0", + "timepoint": 1003, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1003, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1796_0", + "timepoint": 1796, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1796, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,needle,inserting; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,tissue mark,looking; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1032_0", + "timepoint": 1032, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1032, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,body marker,holding; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_1857_0", + "timepoint": 1857, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1857, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,syringe,holding; assistant,needle,holding; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_4_5_1090_0", + "timepoint": 1090, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1090, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,needle,inserting; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,ultrasound probe,looking; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_2127_0", + "timepoint": 2127, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2127, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; cotton,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,disinfection; assistant,forceps,holding; forceps,cotton,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_0_0", + "timepoint": 0, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 0, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_3_4_1748_0", + "timepoint": 1748, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1748, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; needle,patient,touching; assistant,tissue mark,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_3_4_1090_0", + "timepoint": 1090, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1090, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,body marker,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_69_0", + "timepoint": 69, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 69, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; head surgeon,curette,holding " + } + ] + }, + { + "id": "Ultrasound_2_2_1091_0", + "timepoint": 1091, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1091, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,disinfection; assistant,antiseptic,holding; head surgeon,operating table,closeto; assistant,scissors,manipulating; circulator,ultrasound machine,closeto; cotton,patient,touching; scissors,cotton,touching; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_105_0", + "timepoint": 105, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 105, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1071_0", + "timepoint": 1071, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1071, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_832_0", + "timepoint": 832, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 832, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_269_0", + "timepoint": 269, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 269, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "MISS_1_1_1517_0", + "timepoint": 1517, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1517, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,microscope controller,manipulating; anesthetist,anesthesia equipment,controlling; patient,operating table,lyingon; anesthetist,operating table,closeto; anesthetist,patient,anaesthetising; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_302_0", + "timepoint": 302, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 302, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; anesthetist,operating table,manipulating; patient,operating room,entering " + } + ] + }, + { + "id": "MISS_3_6_923_0", + "timepoint": 923, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 923, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,aspirating; assistant,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; anesthetist,anesthesia equipment,controlling; cotton,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_343_0", + "timepoint": 343, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 343, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,patient,aspirating; anesthetist,health monitor,looking; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling; forceps,cotton,touching; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,scalpel,holding; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_983_0", + "timepoint": 983, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 983, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,instrument table,looking; assistant,instrument table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1339_0", + "timepoint": 1339, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1339, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; assistant,antiseptic,holding " + } + ] + }, + { + "id": "MISS_1_1_489_0", + "timepoint": 489, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 489, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,manipulating; patient,operating room,entering; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_44_0", + "timepoint": 44, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 44, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,looking; assistant,instrument table,closeto; head surgeon,ultrasound machine,looking " + } + ] + }, + { + "id": "Ultrasound_3_4_1744_0", + "timepoint": 1744, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1744, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,tissue mark,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; needle,patient,touching; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_2215_0", + "timepoint": 2215, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2215, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; syringe,needle,touching; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,needle,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_817_0", + "timepoint": 817, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 817, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_1772_0", + "timepoint": 1772, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1772, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,needle,holding; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,instrument table,closeto; assistant,syringe,holding; needle,patient,touching; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_548_0", + "timepoint": 548, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 548, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon; circulator,operating table,manipulating " + } + ] + }, + { + "id": "MISS_1_1_772_0", + "timepoint": 772, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 772, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; circulator,microscope,manipulating; patient,operating table,lyingon; circulator,microscope,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_94_0", + "timepoint": 94, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 94, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,antiseptic,looking; circulator,instrument table,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2506_0", + "timepoint": 2506, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2506, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; assistant,syringe,holding; head surgeon,operating table,closeto; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_133_0", + "timepoint": 133, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 133, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "MISS_3_6_653_0", + "timepoint": 653, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 653, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,scalpel,holding; patient,operating table,lyingon; assistant,curette,holding; assistant,instrument table,looking; head surgeon,microscope eye,looking; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; assistant,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2810_0", + "timepoint": 2810, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2810, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; assistant,syringe,injecting; ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_413_0", + "timepoint": 413, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 413, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1099_0", + "timepoint": 1099, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1099, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_918_0", + "timepoint": 918, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 918, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; patient,operating table,lyingon; assistant,body marker,holding; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_889_0", + "timepoint": 889, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 889, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,microscope,manipulating; anesthetist,operating table,closeto; anesthetist,microscope,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_165_0", + "timepoint": 165, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 165, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,herbal disk,removing; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_1_1_1386_0", + "timepoint": 1386, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1386, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope controller,holding; anesthetist,patient,anaesthetising; anesthetist,anesthesia equipment,controlling; assistant,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_84_0", + "timepoint": 84, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 84, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,holding; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_214_0", + "timepoint": 214, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 214, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; forceps,cotton,touching; scalpel,patient,touching; anesthetist,operating table,closeto; head surgeon,scalpel,holding; head surgeon,patient,cutting; anesthetist,health monitor,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,forceps,holding; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "MISS_1_1_1596_0", + "timepoint": 1596, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1596, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,anesthesia equipment,holding; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1266_0", + "timepoint": 1266, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1266, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,ultrasound probe,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,ultrasound screen,looking; assistant,needle,inserting; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_4_5_1269_0", + "timepoint": 1269, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1269, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; assistant,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,ultrasound probe,looking " + } + ] + }, + { + "id": "MISS_3_6_1041_0", + "timepoint": 1041, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1041, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; forceps,cotton,touching; assistant,forceps,holding; patient,operating table,lyingon; head surgeon,microscope eye,looking; cotton,patient,touching; anesthetist,health monitor,looking; head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,patient,aspirating " + } + ] + }, + { + "id": "Ultrasound_3_4_2765_0", + "timepoint": 2765, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2765, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; assistant,needle,removing; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,needle,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; needle,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_319_0", + "timepoint": 319, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 319, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; assistant,forceps,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; forceps,cotton,touching; anesthetist,operating table,closeto; assistant,patient,aspirating; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; cotton,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_3233_0", + "timepoint": 3233, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3233, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,applying; head surgeon,operating table,closeto; tissue paper,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1347_0", + "timepoint": 1347, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1347, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_1115_0", + "timepoint": 1115, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1115, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,holding; circulator,ultrasound machine,closeto; circulator,ultrasound machine,manipulating " + } + ] + }, + { + "id": "MISS_1_1_124_0", + "timepoint": 124, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 124, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_857_0", + "timepoint": 857, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 857, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,antiseptic,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; forceps,cotton,touching; assistant,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "MISS_3_6_488_0", + "timepoint": 488, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 488, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,curette,holding; curette,patient,touching; head surgeon,herbal disk,removing; scalpel,patient,touching; patient,operating table,lyingon; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; head surgeon,scalpel,holding; anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_153_0", + "timepoint": 153, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 153, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,antiseptic,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_466_0", + "timepoint": 466, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 466, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2038_0", + "timepoint": 2038, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2038, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,ultrasound probe,looking; assistant,operating table,closeto; needle,patient,touching; head surgeon,operating table,closeto; assistant,syringe,holding; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; assistant,needle,inserting " + } + ] + }, + { + "id": "Ultrasound_4_5_2517_0", + "timepoint": 2517, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2517, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,tissue paper,holding " + } + ] + }, + { + "id": "MISS_1_1_2014_0", + "timepoint": 2014, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2014, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,patient,disinfection " + } + ] + }, + { + "id": "MISS_1_1_582_0", + "timepoint": 582, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 582, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; patient,operating room,entering; circulator,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_655_0", + "timepoint": 655, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 655, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_530_0", + "timepoint": 530, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 530, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_806_0", + "timepoint": 806, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 806, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; assistant,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "MISS_1_1_1902_0", + "timepoint": 1902, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1902, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; assistant,patient,disinfection; assistant,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,applying; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2593_0", + "timepoint": 2593, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2593, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,needle,holding; head surgeon,ultrasound probe,controlling; syringe,needle,touching; assistant,syringe,injecting; ultrasound probe,patient,touching; assistant,needle,holding; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_5_6_3086_0", + "timepoint": 3086, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3086, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue paper,holding; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,instrument table,closeto " + } + ] + }, + { + "id": "MISS_1_1_841_0", + "timepoint": 841, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 841, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,microscope,manipulating; circulator,microscope,manipulating; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1014_0", + "timepoint": 1014, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1014, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2534_0", + "timepoint": 2534, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2534, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,tissue paper,applying; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_419_0", + "timepoint": 419, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 419, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "Ultrasound_3_4_2310_0", + "timepoint": 2310, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2310, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; assistant,syringe,injecting; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; syringe,needle,touching; head surgeon,needle,holding; needle,patient,touching " + } + ] + }, + { + "id": "MISS_2_2_324_0", + "timepoint": 324, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 324, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; patient,operating table,lyingon; forceps,cotton,touching; cotton,patient,touching; assistant,patient,aspirating; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding; anesthetist,health monitor,looking; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_229_0", + "timepoint": 229, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 229, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating room,entering; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_1571_0", + "timepoint": 1571, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1571, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_1446_0", + "timepoint": 1446, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1446, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,patient,anaesthetising; assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,microscope controller,manipulating " + } + ] + }, + { + "id": "MISS_1_1_279_0", + "timepoint": 279, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 279, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,operating table,manipulating; patient,operating room,entering; anesthetist,operating table,manipulating " + } + ] + }, + { + "id": "MISS_3_6_440_0", + "timepoint": 440, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 440, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,scalpel,holding; head surgeon,microscope eye,looking; assistant,instrument table,closeto; head surgeon,herbal disk,removing; curette,patient,touching; head surgeon,forceps,holding " + } + ] + }, + { + "id": "Ultrasound_3_4_1677_0", + "timepoint": 1677, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1677, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,tissue mark,looking; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,vertebrae,scanning; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2633_0", + "timepoint": 2633, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2633, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; needle,patient,touching; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,syringe,injecting; head surgeon,vertebrae,scanning; assistant,needle,holding; assistant,operating table,closeto; head surgeon,operating table,closeto; syringe,needle,touching " + } + ] + }, + { + "id": "MISS_2_2_175_0", + "timepoint": 175, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 175, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; head surgeon,patient,cutting; patient,operating table,lyingon; scalpel,patient,touching; anesthetist,anesthesia equipment,controlling; assistant,operating table,closeto; head surgeon,scalpel,holding; assistant,forceps,holding; head surgeon,operating table,closeto; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_262_0", + "timepoint": 262, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 262, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,patient,touching; head surgeon,curette,holding; head surgeon,microscope eye,looking; assistant,operating table,closeto; anesthetist,health monitor,looking; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon; curette,patient,touching; head surgeon,operating table,closeto; head surgeon,forceps,holding " + } + ] + }, + { + "id": "MISS_2_2_729_0", + "timepoint": 729, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 729, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,microscope screen,looking " + } + ] + }, + { + "id": "Ultrasound_5_6_2891_0", + "timepoint": 2891, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2891, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,syringe,removing; head surgeon,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,needle,holding; needle,patient,touching; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_2733_0", + "timepoint": 2733, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2733, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; syringe,needle,touching; assistant,syringe,injecting; assistant,needle,holding; needle,patient,touching; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_114_0", + "timepoint": 114, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 114, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instruments,controlling; circulator,syringe,looking " + } + ] + }, + { + "id": "MISS_1_1_2130_0", + "timepoint": 2130, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2130, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; patient,operating table,lyingon; forceps,cotton,touching; assistant,operating table,closeto; cotton,patient,touching; assistant,patient,disinfection; anesthetist,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2125_0", + "timepoint": 2125, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2125, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; needle,patient,touching; syringe,needle,touching; assistant,ultrasound screen,looking; head surgeon,needle,holding; assistant,syringe,injecting; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1345_0", + "timepoint": 1345, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1345, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,body marker,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1961_0", + "timepoint": 1961, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1961, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,patient,disinfection; anesthetist,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; assistant,antiseptic,applying " + } + ] + }, + { + "id": "Ultrasound_5_6_2139_0", + "timepoint": 2139, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2139, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; needle,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; syringe,needle,touching; assistant,syringe,injecting; assistant,needle,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_1350_0", + "timepoint": 1350, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1350, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_373_0", + "timepoint": 373, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 373, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " forceps,cotton,touching; assistant,forceps,holding; head surgeon,scalpel,holding; cotton,patient,touching; assistant,patient,aspirating; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1552_0", + "timepoint": 1552, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1552, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,patient,anaesthetising; head surgeon,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_714_0", + "timepoint": 714, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 714, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,instrument table,closeto; patient,operating table,lyingon; assistant,forceps,holding; head surgeon,patient,manipulating " + } + ] + }, + { + "id": "Ultrasound_5_6_1234_0", + "timepoint": 1234, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1234, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_3_4_1608_0", + "timepoint": 1608, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1608, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; assistant,tissue mark,looking; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_1_1_2226_0", + "timepoint": 2226, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 2226, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; forceps,cotton,touching; assistant,operating table,closeto; cotton,patient,touching; assistant,patient,disinfection " + } + ] + }, + { + "id": "Ultrasound_3_4_2565_0", + "timepoint": 2565, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2565, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,holding; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; assistant,syringe,injecting; head surgeon,needle,holding; head surgeon,vertebrae,scanning; needle,patient,touching; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; ultrasound probe,patient,touching; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_3_4_1610_0", + "timepoint": 1610, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1610, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,operating table,closeto; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,tissue mark,looking; patient,operating table,lyingon; assistant,needle,inserting " + } + ] + }, + { + "id": "MISS_3_6_609_0", + "timepoint": 609, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 609, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " curette,patient,touching; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; assistant,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,operating table,closeto; head surgeon,curette,holding; patient,operating table,lyingon; scalpel,patient,touching; head surgeon,scalpel,holding; anesthetist,health monitor,looking; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "Ultrasound_5_6_1634_0", + "timepoint": 1634, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1634, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,vertebrae,scanning; assistant,instrument table,looking; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_2035_0", + "timepoint": 2035, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2035, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating table,closeto; assistant,needle,inserting; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; ultrasound probe,patient,touching; assistant,ultrasound probe,looking; head surgeon,ultrasound screen,looking; needle,patient,touching " + } + ] + }, + { + "id": "Ultrasound_2_2_368_0", + "timepoint": 368, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 368, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,patient,closeto; patient,operating table,closeto; circulator,operating table,closeto; circulator,patient,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_216_0", + "timepoint": 216, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 216, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,syringe,looking; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2590_0", + "timepoint": 2590, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2590, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; syringe,needle,touching; needle,patient,touching; head surgeon,ultrasound screen,looking; assistant,needle,holding; head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; assistant,syringe,injecting; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_1504_0", + "timepoint": 1504, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1504, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_3_6_729_0", + "timepoint": 729, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 729, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; assistant,instrument table,closeto; assistant,instrument table,looking; assistant,antiseptic,holding; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_5_6_3188_0", + "timepoint": 3188, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3188, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; assistant,tissue paper,applying; patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_735_0", + "timepoint": 735, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 735, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,holding; circulator,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1256_0", + "timepoint": 1256, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1256, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; anesthetist,patient,manipulating; patient,operating table,lyingon; anesthetist,operating table,closeto; assistant,operating table,closeto; anesthetist,anesthesia equipment,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_3265_0", + "timepoint": 3265, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 3265, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " tissue paper,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,tissue paper,applying; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_5_6_221_0", + "timepoint": 221, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 221, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,ultrasound gel,holding; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1140_0", + "timepoint": 1140, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1140, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,forceps,holding; assistant,operating table,closeto; forceps,cotton,touching; assistant,antiseptic,holding; head surgeon,operating table,closeto; circulator,ultrasound machine,manipulating; circulator,ultrasound machine,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_870_0", + "timepoint": 870, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 870, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,forceps,holding; patient,operating table,lyingon; assistant,instrument table,closeto; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_905_0", + "timepoint": 905, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 905, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; assistant,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_1542_0", + "timepoint": 1542, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1542, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_957_0", + "timepoint": 957, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 957, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,anesthesia equipment,controlling; assistant,patient,aspirating; assistant,operating table,closeto; head surgeon,microscope eye,looking; cotton,patient,touching; assistant,forceps,holding; anesthetist,operating table,closeto; head surgeon,operating table,closeto; forceps,cotton,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_1225_0", + "timepoint": 1225, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1225, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,ultrasound machine,closeto; assistant,operating table,closeto; head surgeon,operating table,closeto; assistant,forceps,holding; forceps,cotton,touching; assistant,antiseptic,holding; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_738_0", + "timepoint": 738, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 738, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,microscope eye,looking; head surgeon,operating table,closeto; assistant,instrument table,closeto; anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; assistant,instruments,disinfection; assistant,anesthetist,applying; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2019_0", + "timepoint": 2019, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2019, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound probe,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,syringe,holding; head surgeon,operating table,closeto; needle,patient,touching; ultrasound probe,patient,touching; assistant,needle,inserting; head surgeon,ultrasound screen,looking; head surgeon,vertebrae,scanning; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_134_0", + "timepoint": 134, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 134, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,ultrasound machine,looking; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2615_0", + "timepoint": 2615, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2615, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; needle,patient,touching; assistant,operating table,closeto; assistant,syringe,injecting; assistant,needle,holding; head surgeon,ultrasound screen,looking " + } + ] + }, + { + "id": "Ultrasound_4_5_2585_0", + "timepoint": 2585, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2585, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,tissue paper,applying; patient,operating table,lyingon; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1169_0", + "timepoint": 1169, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1169, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,needle,inserting; assistant,ultrasound probe,looking; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; needle,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_74_0", + "timepoint": 74, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 74, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,ultrasound gel,holding; patient,operating table,lyingon; head surgeon,ultrasound probe,holding; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2205_0", + "timepoint": 2205, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2205, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; needle,patient,touching; syringe,needle,touching; head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_890_0", + "timepoint": 890, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 890, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; circulator,microscope,manipulating; anesthetist,microscope,manipulating; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_2_2_102_0", + "timepoint": 102, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 102, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instruments,controlling; circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_2482_0", + "timepoint": 2482, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2482, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,looking; patient,operating table,lyingon; assistant,instrument table,closeto; head surgeon,ultrasound machine,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_69_0", + "timepoint": 69, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 69, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto; circulator,instrument table,looking " + } + ] + }, + { + "id": "MISS_2_2_579_0", + "timepoint": 579, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 579, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,health monitor,looking; anesthetist,anesthesia equipment,controlling; head surgeon,microscope controller,controlling; anesthetist,operating table,closeto; patient,operating table,lyingon; assistant,operating table,closeto; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_369_0", + "timepoint": 369, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 369, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,ultrasound gel,applying; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; head surgeon,ultrasound probe,holding " + } + ] + }, + { + "id": "Ultrasound_4_5_1156_0", + "timepoint": 1156, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1156, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,vertebrae,scanning; assistant,needle,inserting; head surgeon,operating table,closeto; assistant,ultrasound probe,looking; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_759_0", + "timepoint": 759, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 759, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; anesthetist,anesthesia equipment,holding; anesthetist,operating table,closeto; circulator,microscope,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2931_0", + "timepoint": 2931, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2931, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,tissue paper,applying; head surgeon,operating table,closeto; assistant,operating table,closeto; tissue paper,patient,touching; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_3_6_1048_0", + "timepoint": 1048, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 1048, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " cotton,patient,touching; anesthetist,health monitor,looking; patient,operating table,lyingon; anesthetist,anesthesia equipment,controlling; assistant,forceps,holding; assistant,patient,aspirating; forceps,cotton,touching; assistant,operating table,closeto; anesthetist,operating table,closeto; head surgeon,microscope eye,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2508_0", + "timepoint": 2508, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2508, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " needle,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,ultrasound probe,controlling; assistant,syringe,holding; head surgeon,needle,holding; head surgeon,vertebrae,scanning; patient,operating table,lyingon " + } + ] + }, + { + "id": "Ultrasound_4_5_1063_0", + "timepoint": 1063, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1063, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; ultrasound probe,patient,touching " + } + ] + }, + { + "id": "MISS_3_6_353_0", + "timepoint": 353, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 353, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; curette,patient,touching; patient,operating table,lyingon; head surgeon,forceps,holding; head surgeon,herbal disk,removing; head surgeon,operating table,closeto; head surgeon,curette,holding; assistant,operating table,closeto; head surgeon,microscope eye,looking; forceps,patient,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_1485_0", + "timepoint": 1485, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1485, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,ultrasound screen,looking; ultrasound probe,patient,touching; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2271_0", + "timepoint": 2271, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2271, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; head surgeon,ultrasound probe,controlling; needle,patient,touching; assistant,syringe,injecting; syringe,needle,touching; head surgeon,ultrasound screen,looking; assistant,operating table,closeto; head surgeon,needle,holding; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,needle,holding; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1070_0", + "timepoint": 1070, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1070, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto; ultrasound probe,patient,touching; assistant,operating table,closeto; assistant,needle,holding " + } + ] + }, + { + "id": "MISS_1_1_57_0", + "timepoint": 57, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 57, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,instrument table,closeto; circulator,scalpel,looking " + } + ] + }, + { + "id": "Ultrasound_2_2_542_0", + "timepoint": 542, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 542, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating room,entering; assistant,operating room,entering; patient,operating table,lyingon; circulator,assistant,closeto " + } + ] + }, + { + "id": "MISS_1_1_1347_0", + "timepoint": 1347, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1347, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,instrument table,closeto; patient,operating table,lyingon; anesthetist,patient,anaesthetising; head surgeon,operating table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling; circulator,operating table,looking; assistant,instrument table,looking " + } + ] + }, + { + "id": "MISS_3_6_548_0", + "timepoint": 548, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 548, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,microscope eye,looking; anesthetist,operating table,closeto; assistant,operating table,closeto; curette,patient,touching; head surgeon,curette,holding; scalpel,patient,touching; patient,operating table,lyingon; head surgeon,herbal disk,removing; anesthetist,anesthesia equipment,controlling; head surgeon,scalpel,holding " + } + ] + }, + { + "id": "Ultrasound_5_6_705_0", + "timepoint": 705, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 705, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; patient,operating table,lyingon; assistant,operating table,closeto; ultrasound probe,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_155_0", + "timepoint": 155, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 155, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,ultrasound probe,holding; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_987_0", + "timepoint": 987, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 987, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; circulator,microscope,manipulating; anesthetist,microscope,manipulating; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_5_6_2529_0", + "timepoint": 2529, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 2529, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; patient,operating table,lyingon; needle,patient,touching; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; assistant,needle,holding; head surgeon,ultrasound probe,controlling; ultrasound probe,patient,touching; assistant,syringe,injecting; assistant,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_1094_0", + "timepoint": 1094, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1094, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,scissors,manipulating; patient,operating table,lyingon; assistant,antiseptic,holding; circulator,ultrasound machine,closeto; scissors,cotton,touching; cotton,patient,touching; head surgeon,operating table,closeto; assistant,operating table,closeto; assistant,patient,disinfection " + } + ] + }, + { + "id": "MISS_1_1_6_0", + "timepoint": 6, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 6, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,looking; circulator,operating room,entering " + } + ] + }, + { + "id": "Ultrasound_5_6_1110_0", + "timepoint": 1110, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 1110, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,vertebrae,scanning; assistant,operating table,closeto; ultrasound probe,patient,touching; patient,operating table,lyingon; head surgeon,ultrasound probe,controlling; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_682_0", + "timepoint": 682, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 682, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,patient,manipulating; assistant,operating table,closeto; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_2_2_641_0", + "timepoint": 641, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 641, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,microscope controller,controlling; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,microscope screen,looking; assistant,operating table,closeto; assistant,microscope screen,looking; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "MISS_1_1_1538_0", + "timepoint": 1538, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1538, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " anesthetist,patient,anaesthetising; assistant,operating table,closeto; anesthetist,operating table,closeto; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,microscope controller,manipulating; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_1167_0", + "timepoint": 1167, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 1167, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; assistant,antiseptic,holding; circulator,ultrasound machine,manipulating; patient,operating table,lyingon; forceps,cotton,touching; circulator,ultrasound machine,closeto; assistant,operating table,closeto; assistant,forceps,holding " + } + ] + }, + { + "id": "MISS_3_6_178_0", + "timepoint": 178, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 178, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,forceps,holding; head surgeon,operating table,closeto; head surgeon,curette,holding; anesthetist,anesthesia equipment,controlling; head surgeon,herbal disk,removing; curette,patient,touching; assistant,operating table,closeto; patient,operating table,lyingon; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_2_2_585_0", + "timepoint": 585, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 585, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; head surgeon,operating table,closeto; assistant,operating table,closeto " + } + ] + }, + { + "id": "MISS_2_2_482_0", + "timepoint": 482, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 482, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; patient,operating table,lyingon; assistant,scalpel,holding; anesthetist,health monitor,looking; assistant,instrument table,closeto; anesthetist,operating table,closeto; anesthetist,anesthesia equipment,controlling " + } + ] + }, + { + "id": "Ultrasound_2_2_250_0", + "timepoint": 250, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 2, + "take_id": 2, + "frame_idx": 250, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " circulator,instrument table,closeto " + } + ] + }, + { + "id": "Ultrasound_4_5_1933_0", + "timepoint": 1933, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 1933, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; assistant,operating table,closeto; patient,operating table,lyingon; needle,patient,touching; assistant,syringe,injecting; head surgeon,vertebrae,scanning; syringe,needle,touching " + } + ] + }, + { + "id": "Ultrasound_5_6_499_0", + "timepoint": 499, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 5, + "take_id": 6, + "frame_idx": 499, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; ultrasound probe,patient,touching; head surgeon,vertebrae,scanning; head surgeon,operating table,closeto " + } + ] + }, + { + "id": "Ultrasound_3_4_2295_0", + "timepoint": 2295, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 2295, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound screen,looking; head surgeon,ultrasound probe,controlling; assistant,operating table,closeto; patient,operating table,lyingon; assistant,needle,holding; syringe,needle,touching; needle,patient,touching; assistant,syringe,injecting; ultrasound probe,patient,touching; head surgeon,needle,holding; head surgeon,operating table,closeto; head surgeon,vertebrae,scanning " + } + ] + }, + { + "id": "Ultrasound_4_5_2358_0", + "timepoint": 2358, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 4, + "take_id": 5, + "frame_idx": 2358, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,operating table,closeto; head surgeon,vertebrae,scanning; ultrasound probe,patient,touching; assistant,operating table,closeto; needle,patient,touching; assistant,needle,removing; patient,operating table,lyingon " + } + ] + }, + { + "id": "MISS_1_1_1144_0", + "timepoint": 1144, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 1, + "take_id": 1, + "frame_idx": 1144, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " patient,operating table,lyingon; assistant,operating room,entering; head surgeon,operating room,entering; anesthetist,operating table,closeto " + } + ] + }, + { + "id": "MISS_3_6_529_0", + "timepoint": 529, + "hdf5_indices": { + "surgery_type": "MISS", + "procedure_id": 3, + "take_id": 6, + "frame_idx": 529, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " assistant,operating table,closeto; curette,patient,touching; patient,operating table,lyingon; head surgeon,operating table,closeto; head surgeon,curette,holding; head surgeon,microscope eye,looking; head surgeon,scalpel,holding; scalpel,patient,touching; anesthetist,anesthesia equipment,controlling; anesthetist,operating table,closeto; head surgeon,herbal disk,removing " + } + ] + }, + { + "id": "Ultrasound_3_4_1093_0", + "timepoint": 1093, + "hdf5_indices": { + "surgery_type": "Ultrasound", + "procedure_id": 3, + "take_id": 4, + "frame_idx": 1093, + "available_modalities": [ + "eye_gaze_depth", + "eye_gaze", + "ultrasound", + "exo_frames", + "audio", + "ego_frames", + "hand_tracking", + "point_cloud" + ] + }, + "conversations": [ + { + "from": "human", + "value": "\nEntities: [assistant, anaesthetist, circulator, head surgeon, operating table, operating room, instrument table, antiseptic, anesthesia equipment, bin, body marker, cotton, gloves, health monitor, instruments, microscope controller, microscope eye, microscope screen, needle, patient, scalpel, scissors, forceps, curette, tissue mark, tissue paper, ultrasound gel, ultrasound machine, ultrasound probe, ultrasound screen, unsterile instruments, vertebrae, herbal disk, dressing material]. Predicates: [disinfection, close to, touching, cutting, holding, lying on, manipulating, preparing, scanning, positioning, controlling, looking, inserting, injecting, dropping, applying, dressing, entering, removing, aspirating, wearing, anaesthetising]. Given the following scene graph memory representation, generate a scene graph for timepoint T. The output should strictly be a list of triplets, each in the format \"entity1,entity2,predicate;\". Do not provide a narrative or descriptive text." + }, + { + "from": "gpt", + "value": " head surgeon,ultrasound probe,controlling; head surgeon,vertebrae,scanning; patient,operating table,lyingon; ultrasound probe,patient,touching; assistant,body marker,holding; assistant,operating table,closeto; head surgeon,ultrasound screen,looking; head surgeon,operating table,closeto " + } + ] + } +] \ No newline at end of file