Classification of visual stimuli with different spatial patterns for single-frequency, multi-class SSVEP BCI

Abstract

A new approach for a multi-class steady-state visual-evoked potential (SSVEP)-based brain-computer interface (BCI) is proposed. It was demonstrated through preliminary experiments that spatial patterns of SSVEP responses recorded using high-density electroencephalography while presenting pattern reversal checkerboard stimuli with different spatial patterns can be classified with fairly high accuracy. The average classification accuracies in two of the three subjects were 91.7% (12-class) and 93.3% (15-class), suggesting that the proposed visual stimulation can potentially be used for multi-class SSVEP-based BCI.