Research on auditory BCI based on wavelet transform

It is significant that auditory brain-computer interface (BCI) technology can provide a means of non-muscular communication. For most severely disabled people, limitations in eye mobility or visual acuity may necessitate auditory BCI systems. The auditory BCI based on three-stimulus paradigm was studied to obtain a binary decision in this paper. Coherence average and one-dimensional discrete wavelet transform were used to reduce noise and improved signal-to-noise ratio, and extract P300 feature in low-frequency signals. The target and non-target stimuli were classified by support vector machines. The results show that three-stimulus paradigm could elicit P300 potential and an auditory P300 BCI is feasible. The identification correct rates achieved more than 80%, which can be comparable to the BCI based on visual.