Toward multi-command auditory brain computer interfacing using speech stimuli

Brain-computer interfaces (BCIs) based on event-related potentials (ERP) are promising tools to communicate with patients suffering from some severe disabled diseases. ERP is evoked by various stimuli such as auditory, olfactory, and visual stimuli. Some auditory based BCIs with certain synthetic tone have been proposed, however, it is still challenging to increase the number of commands in auditory-based BCIs, since it is usually difficult for users to remember and distinguish multiple tones that corresponds to commands. We propose a new auditory BCI framework using speech stimuli. It is easier for users to distinguish different speech stimuli than different simple tones. We show experimental results of four-command BCI. The proposed speech-based BCI achieved a classification accuracy of more than 70 percents.