Recognition and Classification of Spatial Auditory Evoked P300 EEG Signal

Auditory brain computer interface technology provides a new way of communicating for people with blocked visual pathways. The experimental paradigm based on spatial auditory evoked P300 was designed here. 9 subjects kept count of target stimuli under the stimuli of three-tone oddball paradigm with spatial sound properties. Original P300 signal was processed by wavelet analysis and interpolation, self-correlation denoising of empirical mode decomposition based on noise statistics, and less times superposition and averaging. An electrode selected algorithm based on independent component analysis was adopted. The features in time domain of P300 at selected electrodes were extracted and classified by stepwise linear discriminant analysis, and the consistency of the classification was evaluated. The best recognition rate of 9 subjects was 80.2%, and the average recognition rate was 74.4%, which verified the feasibility of the spatial auditory evoked P300 experiment and the effectiveness of the algorithm.