RESUMO
The goal of this work was to describe a system for real-time typing controlled by brain biopotential signals. A 6 ( 6 matrix containing Russian alphabet letters and auxiliary symbols was shown on PC screen. Electroencephalogram was taken, and the P300 component was extracted (this component appeared only upon presentation of a significant stimulus). A combination of several detection methods was used to identify the P300 component, which made it possible to increase the probability of correct identification to 91.6 (5.2%. It was shown that the developed interface could be implemented on the basis of a single active electrode in Pz (Cz) position.
Assuntos
Computadores , Eletroencefalografia/métodos , Potenciais Evocados P300 , Software , Algoritmos , HumanosRESUMO
In four adult healthy subjects in 18 experiments, we studied Brain-Computer-Interface recognition of different intended words by P300 wave in the VEP. The set of optimal characteristics of visual stimulation which rise reliability of recognition up to 100 %, as well as effective registration locus (Pz) were determined. It was found that the best processing criteria for letter recognition were: P300 square and superposition of all three criteria (P300 amplitude, square and covariation coefficient).