motor-imagery-BCI

A MATLAB toolbox for classification of motor imagery tasks in EEG-based BCI system with CSP and FB-CSP

Codes and data for the following paper are extended to different methods:

1# Iterative Subspace Decomposition for Ocular Artifact Removal from EEG Recordings

2# Filter bank common spatial pattern algorithm on BCI competition IV Datasets 2a and 2b

1. Introduction.

This package includes the prototype MATLAB codes for motor imagery brain-computer interfaces.

The implemented method uses CSP for feature extraction and SVM as the classifier:

1. Iterative Subspace Decomposition for Ocular Artifact Removal
2. common spatial pattern algorithm to extract features
3. SVM and Naive-Bayes classification

2. Usage & Dependency.

Dependency:

 https://github.com/hisuk/MI-BSSFO     
 https://github.com/alexandrebarachant/covariancetoolbox
 EEGLAB tolbox
 Biosig plugin for eeglab

Usage:

Run "main_pw.m" or "main_ovr.m" to perform A-Z for MI-BCI.