Brain–computer interface for single-trial EEG classification for wrist movement imagery using spatial filtering in the gamma band

Abstract

The aim of this study is to discriminate between the left and right wrist movements imagery in four different directions. To achieve this goal, the authors have applied spatial filtering on the EEG signal in the gamma frequency band to extract key features to perform classification. Specifically, the original EEG signal is transformed in to a spatial pattern and applied to the radial basis function (RBF) classifier. The authors demonstrate that spatial filtering method in multichannel EEG effectively extracts discriminant information from single-trial EEG for left and right wrist movement imagery. An average recognition rate of approximately 89% was achieved in all the four type of movements (extension, flexion, pronation and supination) between left and right wrist in five healthy subjects. The results are comparable to the highest rates reported in the literature.