Noise-assisted auditory brain computer interface

説明

Recently, brain computer interface (BCI) has been extensively studied as a support tool for severe physical disabilities. In particular, BCIs using visual stimuli (Visual-BCIs) have been successfully developed in the speed and accuracy. However, there exists a crucial problem that visual impaired people cannot use the Visual-BCIs. It is thus useful to develop eye-independent BCIs which use other modalities rather than visual stimuli. We focus on an auditory BCI using modulated sound stimuli which elicit auditory steady-state response (ASSR) as one of the eye-independent BCIs. Conventional auditory BCIs based on ASSR (ASSR-BCI) have had low performance due to small amplitude of ASSR. The present study thus aimed at improvement of amplitude and performance of ASSR-BCI using stochastic resonance. As a result, classification accuracy of the proposed BCI was 77.2% under an additive noise condition with a support vector machine, which was improved by 10.5% compared to conventional ASSR-BCI under noiseless condition.

収録刊行物

詳細情報 詳細情報について

問題の指摘

ページトップへ