Mining multi-channel EEG for its information content: an ANN-based method for a brain-computer interface.

We have studied 56-channel electroencephalograms (EEG) from three subjects who planned and performed three kinds of movements, left and right index finger, and right foot movement. Using autoregressive modeling of EEG time series and artificial neural nets (ANN), we have developed a classifier that can tell which movement is performed from a segment of the EEG signal from a single trial. The classifier's rate of recognition of EEGs not seen before was 92-99% on the basis of a 1s segment per trial. The recognition rate provides a pragmatic measure of the information content of the EEG signal. This high recognition rate makes the classifier suitable for a so-called 'Brain-Computer Interface', a system that allows one to control a computer, or another device, with ones brain waves. Our classifier Laplace filters the EEG spatially, but makes use of its entire frequency range, and automatically locates regions of relevant activity on the skull.