'P300 SPELLER' BRAIN-COMPUTER INTERFACE: ENHANCEMENT OF P300EVOKED POTENTIAL BY SPATIAL FILTERS

摘要

Brain-Computer Interfaces (BCI) are communication systemsthat use brain activity to control a computer or otherdevices. The BCI system described in this study is basedon the P300 speller BCI paradigm designed by Farwell anDonchin in 1988 [1]. A new unsupervised algorithm is proposedin this paper1. It is based on the projection of the rawEEG signal into the estimation of the P300 subspace. In thisalgorithm, brain responses to the target stimuli and to the nontarget stimuli are taken into account. They provide a betterestimation of the P300 subspace main components. Datarecorded on three subjects were used to evaluate the proposedmethod. The results are presented using a Bayesian lineardiscriminant analysis (BLDA) classifier.