Classi cation of Electro-encephalographic Spatial Patterns

Abstract The aim of this study is to describe a general approach to determine important electrode positions in the case when the measured EEG-signal is used for classication. To classify planning of movement of right and left index nger, three dierent approaches were compared: classi cation using a physiologically motivated set of four electrodes, a set determined by principal component analysis and electrodes deter-mined by spatial pattern analysis. Spatial pattern analysis enhanced the classi cation rate signicantly from 61:3 1:8 % (with four electrodes) to 71:8 1:4 % whereas the classication rate using the principal component analysis is signicantly lower (65:2 1:4%). Most of the 61 electrodes used had no influence on the classication rate so that in future experi-ments the setup can be simplied drastically to 6 to 8 electrodes without loss of information.