Detecting and tracking multiple directional movements in EEG based BCI

Abstract

Detection of multiple directional movements could be useful in designing a BCI based upper-limb rehabilitation system for stroke patients. Under the experiment protocol of voluntary right-hand center-out movement in four orthogonal directions, we will discuss how to classify the movement directions and speeds in the spatial-temporal-spectra domains by utilizing regularized wavelet-common spatial pattern, mutual information-based feature selection, adaptive trajectory tracking, and source localization.