EEG Signal Processing to Control a Finger Rehabilitation System

Abstract

This study aims to provide a comprehensive comparison for classification of Electroencephalography (EEG) signal based motor imagery, in time domain and time-frequency domain with different classifiers. We used EEG signals recorded while the subjects were imagining the movement of individual fingers, and analyzed the signals in time domain as well as using wavelet transform and Wigner transform. Our main goal is to compare different methods of feature extraction and classification as the important steps in the process of EEG signals for the Brain-Computer Interface (BCI) system. The experimental results indicate that the Support Vector Machine (SVM) method provides a better classification performance compared with other classification methods. Also, Linear Discriminative Analysis (LDA) performs as well as the SVM, after applying PCA for dimension reduction. The proposed scheme can be applied successfully to BCI systems where the amount of large data.