metadata

task_categories:
  - question-answering
language:
  - en
tags:
  - medical
pretty_name: Med_data
size_categories:
  - 100K<n<1M

Complete Dataset

Data shown below is complete Medical dataset

Access the complete dataset using the link below:

Download Dataset

Support Us on Product Hunt and X!

| | |

Connect with Me on Happenstance

Join me on Happenstance!
Click here to add me as a friend

Looking forward to connecting!

For more information or assistance, feel free to contact us at [email protected].

short_description: Medical datasets for healthcare model training.

Medical Datasets

This Medical dataset is crafted as a versatile resource for enthusiasts of data science, machine learning, and data analysis. It replicates the characteristics of real-world healthcare data, offering users a platform to practice, refine, and showcase their data manipulation and analytical skills within the healthcare domain.

Medical Datasets for AI/ML Models

This repository contains a collection of high-quality medical datasets designed for training and fine-tuning AI/ML models in healthcare applications. These datasets cover various aspects of medical knowledge, from question-answering to patient records and specialized medical content.

Dataset Overview

File Name	Description
GenMedGPT-5k.json	General Medical GPT 5k - A comprehensive collection of 5,000 medical instruction-response pairs, covering diverse medical topics and designed to train language models in responding to general medical queries with accurate and detailed information.
HealthCareMagic-100k.json	HealthCare Magic 100k - An extensive dataset containing 100,000 doctor-patient interactions from healthcare platforms, featuring real medical consultations with diagnoses, treatment recommendations, and medical advice across various specialties.
evaluation-medical-instruction-dataset.json	Evaluation Medical Instruction Dataset - A specialized collection designed to evaluate the performance of medical AI models on following complex clinical instructions, featuring challenging cases that test accurate diagnosis and treatment recommendation capabilities.
general-medical-instruction-dataset.json	General Medical Instruction Dataset - A broad-spectrum dataset containing diverse medical instructions and corresponding expert responses covering general medicine, specialties, procedures, and patient education materials for comprehensive medical AI training.
medical-preference-data.json	Medical Preference Data - A dataset capturing healthcare preferences and decision-making patterns from both patients and healthcare providers, useful for training AI systems to align with human preferences in medical contexts.
medical_meadow_wikidoc.json	Medical Meadow WikiDoc - A structured collection of medical knowledge extracted from WikiDoc medical encyclopedia, containing comprehensive information on diseases, treatments, procedures, and medical terminology in a format optimized for AI learning.
medical_meadow_wikidoc_medical_flashcards.json	Medical Meadow WikiDoc Medical Flashcards - A specialized educational dataset featuring concise question-answer pairs on key medical concepts, designed in a flashcard format to train AI models for medical education applications.
medical_meadow_wikidoc_patient_info.json	Medical Meadow WikiDoc Patient Information - A patient-focused dataset containing simplified explanations of medical conditions, treatment options, and healthcare procedures written in accessible language for patient education purposes.
medicationqa.json	Medication Question and Answer - A targeted dataset containing 2,762 medication-related queries and expert responses covering drug dosages, side effects, interactions, administration methods, and pharmacological properties to train AI in medication guidance.
medmcqa-train-instruction-dataset.json	Medical Multiple Choice Question and Answer Training Instruction Dataset - A comprehensive collection of medical multiple-choice questions with detailed explanations for correct answers, covering various medical disciplines and designed for training diagnostic reasoning.
medqa-train-instruction-dataset.json	Medical Question and Answer Training Instruction Dataset - A specialized dataset featuring complex medical questions paired with detailed expert answers, designed to train AI models in providing evidence-based responses to clinical queries.
open-platypus.json	Open Platypus Medical Dataset - An open-source collection of diverse medical content including clinical notes, research findings, and medical literature excerpts, designed to improve AI understanding of medical terminology and concepts.
pubmedqa-train-instruction-train.json	PubMed Question and Answer Training Instruction Dataset - A research-oriented dataset derived from PubMed scientific literature, featuring biomedical research questions with evidence-based answers supported by published medical literature.
umls.json	Unified Medical Language System (UMLS) Dataset - A comprehensive collection of medical terminology, concepts, and coding systems that provides standardized vocabulary for medical AI, ensuring consistent understanding of medical terms across different contexts.
umls_relation.json	Unified Medical Language System (UMLS) Relation Dataset - A specialized dataset defining relationships between medical concepts, terms, and entities within the UMLS framework, enabling AI systems to understand complex connections in medical knowledge.

Detailed Dataset Descriptions

Medication Question and Answer (medicationqa.json)

This dataset contains 2,762 medication-related questions and answers in a structured format:

instruction: Prompt instructing an AI to act as a medical professional
input: Patient/user query about medications (e.g., dosages, side effects, usage instructions)
output: Detailed professional response to the medication question

This dataset is valuable for training models that can:

Answer medication-related questions accurately
Provide dosage information and usage guidelines
Explain side effects and drug interactions
Advise on medication storage and administration

General Medical GPT 5k (GenMedGPT-5k.json)

A carefully curated collection of 5,000 medical instruction-response pairs covering a wide range of medical topics including diagnoses, treatments, procedures, and medical explanations. The dataset is structured to help language models learn appropriate medical responses with proper medical terminology and accurate information.

HealthCare Magic 100k (HealthCareMagic-100k.json)

This extensive dataset contains 100,000 real doctor-patient interactions collected from healthcare platforms. It includes patient queries, symptoms descriptions, doctor responses, diagnoses, and treatment recommendations across various medical specialties, providing rich training material for medical conversational AI.

Evaluation Medical Instruction Dataset (evaluation-medical-instruction-dataset.json)

Specifically designed to evaluate medical AI models, this dataset contains challenging medical instructions that test a model's ability to follow complex clinical directives. It covers edge cases and uncommon scenarios that require deep medical knowledge and careful reasoning.

General Medical Instruction Dataset (general-medical-instruction-dataset.json)

A comprehensive collection of medical instructions and expert responses covering general medicine, specialized fields, diagnostic procedures, treatment protocols, and patient education. This dataset is ideal for training broad-coverage medical AI assistants.

Medical Preference Data (medical-preference-data.json)

This dataset captures human preferences in medical decision-making, including comparisons between different AI responses to medical queries. It can be used to train AI systems through preference learning, aligning AI outputs with human judgments of quality, accuracy, and helpfulness in medical contexts.

Medical Meadow WikiDoc (medical_meadow_wikidoc.json)

A structured repository of medical knowledge extracted from WikiDoc, containing comprehensive information on diseases, conditions, treatments, and medical terminology. The content is organized to facilitate AI learning of medical concepts and relationships.

Medical Meadow WikiDoc Medical Flashcards (medical_meadow_wikidoc_medical_flashcards.json)

This educational dataset features concise question-answer pairs on key medical concepts in a flashcard format. It covers anatomy, physiology, pathology, pharmacology, and clinical procedures, making it ideal for training AI systems for medical education.

Medical Meadow WikiDoc Patient Information (medical_meadow_wikidoc_patient_info.json)

Focused on patient education, this dataset contains simplified explanations of medical conditions, treatment options, and healthcare procedures written in accessible language. It helps train AI to communicate complex medical information to non-specialists.

Medical Multiple Choice Question and Answer Training Instruction Dataset (medmcqa-train-instruction-dataset.json)

A comprehensive collection of medical multiple-choice questions with detailed explanations for correct and incorrect answers. It covers various medical disciplines and is designed to train AI in diagnostic reasoning and medical knowledge assessment.

Medical Question and Answer Training Instruction Dataset (medqa-train-instruction-dataset.json)

This dataset features complex medical questions paired with detailed expert answers, covering clinical scenarios, diagnostic challenges, and treatment decisions. It's structured to train AI in providing evidence-based responses to sophisticated clinical queries.

Open Platypus Medical Dataset (open-platypus.json)

An open-source collection of diverse medical content including clinical notes, research findings, and literature excerpts. This dataset helps improve AI understanding of medical terminology, concepts, and reasoning across different medical contexts.

PubMed Question and Answer Training Instruction Dataset (pubmedqa-train-instruction-train.json)

A research-oriented dataset derived from PubMed scientific literature, featuring biomedical research questions with evidence-based answers. It helps train AI to understand, interpret, and respond to queries about medical research findings.

Unified Medical Language System Dataset (umls.json)

This dataset provides a comprehensive collection of medical terminology, concepts, and coding systems from the Unified Medical Language System. It ensures AI systems have access to standardized vocabulary and can accurately interpret diverse medical terminology.

Unified Medical Language System Relation Dataset (umls_relation.json)

This specialized dataset defines relationships between medical concepts, terms, and entities within the UMLS framework. It enables AI systems to understand the complex connections between different medical concepts, enhancing their ability to reason about medical knowledge.

Healthcare Applications

These datasets can be used to develop AI/ML models that:

Clinical Decision Support Systems: Assist healthcare providers with evidence-based recommendations
Medical Q&A Systems: Answer patient and provider questions about conditions, treatments, and medications
Medical Education Tools: Create intelligent tutoring systems and training simulations
Patient Information Systems: Generate accurate, accessible information for patients
Medication Management: Provide guidance on drug usage, interactions, and side effects
Medical Documentation: Aid in creating and processing medical records and documentation
Diagnostic Support: Help analyze symptoms and suggest possible diagnoses for further investigation

Benefits to Healthcare

Implementing AI/ML models trained on these datasets can provide numerous benefits:

Improved Accessibility: Make medical knowledge more accessible to patients and healthcare providers
Enhanced Efficiency: Reduce time spent on routine questions and information retrieval
Error Reduction: Decrease medication errors through accurate information delivery
Standardized Knowledge: Provide consistent, evidence-based information
Continuous Learning: Models can be updated with new medical discoveries and guidelines
Personalized Care: Enable more tailored responses to individual patient needs
Global Reach: Extend medical expertise to underserved regions and populations

Usage Considerations

When using these datasets for AI/ML model development:

Data Privacy: While these datasets appear to be de-identified, always ensure compliance with healthcare data regulations
Medical Accuracy: Validate model outputs against current medical standards before deployment
Ethical AI: Consider ethical implications of AI-generated medical advice and information
Model Limitations: Clearly communicate the limitations of AI models to end-users
Expert Oversight: Maintain healthcare professional oversight for AI-generated medical content

Citation

When using these datasets in your research or applications, please provide appropriate citations to the original data sources.

Usage

These datasets can be used to train and evaluate models for medical question answering and other related tasks. Each dataset is provided in JSON and TXT formats and contains various medical instructions, inputs, and outputs.

Example

Here is an example entry from the medicationqa.json dataset:

{
    "instruction": "Your role as a doctor requires you to answer the medical questions taking into account the patient's description.",
    "input": "how fast does the medication take to work amlodipine",
    "output": "After oral administration of therapeutic doses of amlodipine besylate, absorption produces peak plasma concentrations between 6 and 12 hours."
}