Designing optimal spatial filters for single-trial EEG classification in amovement task

We devised spatial filters for multi-channel EEG that lead to signals which discriminate optimally between two conditions. We demonstrate the effectiv eness of this method by classifying single-trial EEGs, recorded during prep aration for movements of the left or right index finger or the right foot. The classification rates for 3 subjects were 94, 90 and 84%, respectively. The filters are estimated from a set of multichannel EEG data by the method of Common Spatial Patterns, and reflect the selective activation of cortic al areas. By construction, we obtain an automatic weighting of electrodes a ccording to their importance for the classification task. Computationally, this method is parallel by nature, and demands only the evaluation of scala r products. Therefore, it is well suited for on-line data processing. The r ecognition rates obtained with this relatively simple method are as good as , or higher than those obtained previously with other methods. The high rec ognition rates and the method' s procedural and computational simplicity ma ke it a particularly promising method for an EEG-based brain-computer inter face. (C) 1999 Elsevier Science Ireland Ltd. All rights reserved.