Comparative study between subband and standard ICA/BSS method in context with EEG signal for movement imagery classification

Abstract

This paper work exploits the effectiveness of subband Independent component analysis(ICA)/blind source separation (BSS) in context with EEG signals over the standard ICA/BSS method. The estimated separating matrix by both methods is further subjected to the EOG corrected EEG signals for the extraction of the temporally decorrelated EEG signals. We propose the novel method for automatic selection of the temporally decorrelated /independent components, which have maximum discriminatory information (that captures the phenomenon of ERD and ERS) among the signal subspace components of signal space. The performance of the proposed method has been evaluated by classification accuracy and Cohen's kappa coefficient (k).