BIGP3B Dataset EEG BCI Analysis

This repository hosts a Jupyter notebook for loading, preprocessing, and analyzing the BigP3BCI dataset, focusing on P300 oddball event‑related potentials (ERPs) in an EEG brain–computer interface (BCI) paradigm.

Contents

• BIGP3BDataset_EEG_BCI.ipynb – The main analysis notebook:

  • Reading and filtering raw EDF files
  • Channel renaming and montage setting
  • Event detection and onset index computation
  • Epoching target vs. non‑target trials
  • ERP averaging and comparison at midline electrodes (Fz, Cz, Pz)
  • Feature extraction via windowed means and classifier pipeline

Prerequisites

• Python ≥ 3.10
• Conda for environment management

Installation

1. Clone this repository:

   git clone https://github.com/yourusername/BIGP3B_EEG_BCI.git
   cd BIGP3B_EEG_BCI

2. Create and activate a Conda environment:

   conda create -n bci_env python=3.10 -y
   conda activate bci_env

3. Install required packages:

   conda install -c conda-forge mne numpy scipy pandas matplotlib jupyter -y

Usage

1. Download the BigP3BCI data (e.g. from PhysioNet):

   # example: place `C_04_SE001_CB_Train02.edf` and related files in `data/`

2. Launch Jupyter Lab or Notebook:

   jupyter lab

3. Open and run BIGP3BDataset_EEG_BCI.ipynb:

   • Step through cells to preprocess, epoch, and visualize P300 ERPs.
   • Modify parameters (e.g., epoch window, channels) as needed.

Data Description

The BigP3BCI dataset contains single‑trial EEG recordings during an auditory oddball task. Key channels:

• StimulusBegin – Square pulse indicating flash onset
• StimulusType – Code for target vs. non‑target
• PhaseInSequence – Position in the stimulus sequence

Refer to the dataset documentation for full details: https://physionet.org/content/bigp3bci/1.0.0/

License

This project is licensed under the MIT License. See LICENSE for details.

---

© 2025 Andy Gibson