Brain Computer Interface Using Modulation of Auditory Steady-State Response with Help of Stochastic Resonance.

This paper proposes an eye-movement independent brain computer interface based on the modulations of auditory steady-state response (ASSR-BCI) to amplitude-modulated (AM) tones elicited by paying selective attention to one of the two AM tones. Moreover, the proposed ASSR-BCI exploits a stochastic resonance effect to improve the signal separation and attained the mean classification accuracy of 77 % across nine normal subjects under a noise-added condition with sound pressures 60 dB for the two tones and 30 dB for the noise added to the two AM tones. Results from information transfer rate and its inter-individual difference suggest that it may be adequate to set an inter-trial interval at 2∽3 s for a trial time length. It is consequently feasible to develop a practical eye-movement-independent BCI available in eyes-closed state by optimizing the parameters such as the trial time length and electrode sites each user.