Goal-Directed Learning in Cortical Organoids - Experimental Data and Code

This repository contains the experimental data and analysis framework supporting the paper "Goal-Directed Learning in Cortical Organoids."

Contents

Experimental Data

• aggregated_experiment_data.pkl - Processed HD-MEA recordings, neural unit activity, task trajectories, performance metrics, and connectivity matrices from 19 cortical organoids across 432 training cycles
• README_data_structure.md - Detailed documentation of data structure, file formats, and experimental metadata
• data_access_example.py - Example script demonstrating data loading and basic analysis

Code (BrainDance Framework)

The BrainDance framework (https://github.com/braingeneers/brainDance) provides the experimental infrastructure for closed-loop organoid training.

Experimental phases (run individually):

• recording_exp_def.py - Spontaneous recording phase for neural unit identification
• causal_exp_def.py - Stimulus-response characterization phase
• cartpole_exp_def.py - Closed-loop training phase with cartpole task

Core infrastructure:

• maxwell_env.py - Hardware interface for MaxWell HD-MEA system
• phases.py - Phase definitions for experimental workflow
• phases_analysis.py - Real-time and post-hoc analysis tools
• causal_connectivity.py - Causal connectivity analysis and artifact removal
• data_loader.py - Data loading utilities
• trainer.py - Reinforcement learning implementation for adaptive training

Getting Started

1. See README_data_structure.md for detailed data organization
2. Run data_access_example.py to explore the data
3. Visit https://braingeneers.github.io/braindance for full framework documentation

Citation

If you use this data or code, please cite:

Robbins, A., et al. (2025). Goal-Directed Learning in Cortical 
Organoids - Experimental Data and Code. Zenodo. 
https://doi.org/10.5281/zenodo.17684862