A domotic control system using Brain-Computer Interface (BCI)

Abstract

The aim of this study was to develop a Brain Computer Interface (BCI) application to control domotic devices usually present at home. Electroencephalographic (EEG) activity was recorded from users' scalp and sensorimotor rhythms were used to control the BCI. Our application uses the BCI2000 general purpose system. We studied four feature extraction algorithms: AR model, Fourier transform, wavelet transform and matched filter. An AR model achieved the best accuracy. Our application allows to control several devices, such as a TV set, a Hi-Fi system, a DVD player and the lights of a room. Five users from a disability reference center in León (Spain) with severe physical and cognitive disabilities took part in the study, achieving accuracies up to 69.9%. Domotic BCI applications could be really useful for disabled people. However, further research is needed. © 2011 Springer-Verlag.