1st order class separability using EEG-based features for classification of wrist movements with direction selectivity

28 channel EEG data were recorded while a subject performed wrist movements in four directions. Four feature types were extracted for each channel following optimized filtering of the signals. The potential performance of each feature and channel for use in the classification of the EEG signals was analyzed by estimating the relative class overlap using a first order histogram approach. The best feature/channel configurations contained channels both that were close and far from motor areas. While the scope and depth of the study was very limited, the results do suggest more attention should be paid to non-motor areas when investigating movement related EEG.