Submission for Clinical BCI Challenge
Team leader is Marie-Constance Corsi (Aramis project-team, Inria Paris, Paris Brain Institute, Paris, France), assisted by Florian Yger (LAMSADE, Univ. Paris-Dauphine, Paris, France) and Sylvain Chevallier (LISV, Univ. Paris-Saclay, Velizy, France).
Matlab files are first run to extract features from raw dataset using the Brainstorm toolbox, saving results in .mat files. The .mat files are opened in Python to make the prediction.
- Download the dataset from GitHub
- Run "ComputeFCEstimators.m" matlab file to extract the features (in /Matlab/Matlab_code)
- Run "python rigoletto-predict.py"
The extracted features are available in /Matlab/Matlab_db in .mat files (cf the associated subfolders) or via the Brainstorm project ("RIGOLETTO_bst_project").
The necessary packages for running the python script are indicated in the requirements.txt file and could be installed with pip install -r requirements.txt.
The code is released under GNU GPL v3