An EEG-based brain—computer interface with real-time artifact removal using independent component analysis

This paper presents a steady-state visually evoked potential (SSEVP) EEG-based brain-computer interface (BCI) with real-time artifact removal using independent component analysis (ICA). This system comprises an SSEVP-based stimulator, a four-channel electroencephalogram (EEG) front-end module, an ICA-based artifact removal module, and a canonical correlation analysis (CCA) detection module. The proposed SSEVP-based BCI system with artifact removal module is implemented and verified using ARM-based system on chip platform. The test patterns of 9Hz and 10Hz SSEVP raw data with a 2-second window, advancing with a 1-sec interval, are applied to this system, and the improvements of the hit rate are achieved approximately 53% and 48%, respectively.