continuous_control_bci

An analysis of EEG data for continuous real-time BCI control.

Installation

Install pipenv with pip install pipenv

Then create a virtual environment with pipenv install

Run instructions

Before you can run any analysis you should prepare the ICA decomposition with pipenv run make_ica.py

You can run the different analysis scripts with pipenv run some_script.py

The scripts are generally separated into a calibration version and a driving version to run the analysis on the calibration or driving data. The data that this all corresponds is not yet publicly available, but may be made available on request.

ICA instructions

With 20 participants in 2 contexts there are 40 sessions for which ICs need to be rejected. The ICA decomposition can all be generated with make_ica.py. Then you can inspect each decomposition with show_ica.py. Any ICs marked as bad during this inspection will be stored as bad. Subsequent analyses will run with those ICs are rejected.

Notebooks

There are several Jupyter notebooks in /notebooks. These are kept for posteriority and may be considered useful, but they are not intended to be reproduced.

Questions

Any questions may be directed to ivo.de.jong@rug.nl.