Classification of evoked potentials by Pearson's correlation in a brain-computer interface

François Cabestaing (1) , Theresa M Vaughan (2) , Dennis J Mcfarland (2) , Jonathan R Wolpaw (2)

1 LAGIS-SI
2 Laboratory of Neural Injury and Repair

In this paper, we describe and evaluate the performance of a linear classifier learning technique for use in a brain-computer interface. Electroencephalogram (EEG) signals acquired from individual subjets are analyzed with this technique in order to detect responses to visual stimuli. Signal processing and classification are used for implementing a palliative communication system which allows the individual to spell words. Performance with this technique is evaluated on data collected from eight individuals.

Mots clés: Brain-computer interface; BCI; linear classifier; evoked potentials; P300

Domaines: Traitement du signal et de l'image [eess.SP]; Traitement du signal et de l'image [eess.SP]

François Cabestaing, Theresa M Vaughan, Dennis J Mcfarland, Jonathan R Wolpaw. Classification of evoked potentials by Pearson's correlation in a brain-computer interface. AMSE, Modelling C Automatic Control (theory and applications), 2007, 67, pp.156-166. ⟨hal-00521049⟩