geseft/rat_cortical_128ch

🔬 High-Resolution Cortical Activity Dataset

Electrophysiological recordings from anesthetized rats

📖 Table of Contents

1. 🚀 Overview
2. 🔗 Original Source
3. 📝 Description
4. 📂 Repository Structure
5. 🔧 NWB File Structure
6. 📚 Related Publications
7. 📥 Download
8. 📧 Contact

---

🚀 Overview

This dataset provides both raw and preprocessed wideband recordings from the neocortex of 20 ketamine/xylazine-anesthetized rats, captured with a 128-channel silicon probe (32×4 layout).

• Total size: ~0.9 TB
• Recordings: 109 spontaneous sessions
• Single units: 7,126 sorted
• Format: NWB:N 2.0

---

🔗 Original Source

• Mirror 1
• Mirror 2

---

📝 Description

Summary

• High‐density recordings: 128 titanium-nitride microelectrodes, ultralow impedance
• Waveform data: Raw & filtered spike waveforms; mean & SD across channels
• Quality metrics: SNR, RMS noise/signal, 50 Hz line noise contamination
• Metadata: Animal weight, sex, cortical area & stereotaxic coordinates

All sessions are packaged as individual NWB files, ready for analysis in Python, MATLAB, or any NWB-compatible tool.

Summary

Here, we present an electrophysiological dataset (~0.9 TB overall size) recorded from the neocortex of twenty rats anesthetized with ketamine/xylazine (Horváth et al., 2021). The wideband, spontaneous recordings (n = 109) were acquired with a single-shank silicon-based probe having 128 densely packed recording sites arranged in a 32x4 array (Fiáth et al., 2018). The dataset contains the activity of a total of 7126 sorted single units extracted from all layers of the cortex. Here, we share raw neural recordings, as well as spike times, extracellular spike waveforms and several properties of units packaged in a standardized electrophysiological data format. For technical validation of our dataset, we provide the distributions of derived single unit properties along with various spike sorting quality metrics.

---

📂 Repository Structure

├── raw_data/
│   ├── Rat(01-20)_Insertion(1-2)_Depth(1-3).nwb
│   ├── allSingleUnits.nwb

• Per-animal folders: Organized by Rat ID → Insertion → Depth
• CSV files:
  • Animal_characteristics_and_targeted_cortical_areas.csv: weights, sex, cortical targets, coordinates
  • Recording_characteristics.csv: file size, duration, unit yield, noise & SNR metrics

---

🔧 NWB File Structure

Each .nwb file contains:

1. acquisition/

   • wideband_multichannel_recording (raw continuous data)
   • Units & conversion factors

2. general/

   • devices & extracellular_ephys: probe details, electrode geometry
   • subject: animal metadata (ID, species, sex, weight)

3. units/

   • Spike times & counts
   • Mean & SD waveforms (filtered & raw) for all 128 channels
   • 32×4 layout arrays for easy 2D visualization
   • Peak-channel waveforms
   • Session ID & cortical area annotations

🔍 Explore in HDFView.

---

📚 Related Publications

• Fiáth et al. (2018)
  A silicon-based neural probe with densely-packed low-impedance titanium nitride microelectrodes for ultrahigh-resolution in vivo recordings.
  Biosensors and Bioelectronics, 106, 86–92.
  DOI: https://doi.org/10.1016/j.bios.2018.01.060

• Horváth et al. (2021)
  Dataset of cortical activity recorded with high spatial resolution from anesthetized rats.
  Scientific Data, 8, 180.
  DOI: https://doi.org/10.1038/s41597-021-00970-3

---

📥 Download

git clone https://huggingface.co/datasets/geseft/rat_cortical_128ch

Or download individual files via the Hugging Face web interface.

---

📧 Contact

Have questions or feedback? Please open an issue.

---

© 2025. Licensed under CC BY 4.0.