Effects of the combination of block-shape and face flashing stimuli on a P300 brain computer interface

Abstract

Brain-computer interface (BCI) provides a new way for the brain to interact with the outside world. One of the common BCI systems is the visual P300 speller. Previous studies have shown that faces can induce strong event-related potentials (ERPs) to achieve high classification accuracy and the block-shaped condition can obtain better performance by increasing the size of the highlighted surface. However, the effects of both blended factors have not yet been studied. Therefore, we proposed a red block-shaped face paradigm (RBFP) combining both block-shape and face. We compared the proposed paradigm with the traditional red face paradigm (RFP) to validate its feasibility and effectiveness. Results indicated that RBFP had better classification accuracy (95.27% vs. 91.11%) and raw bit rate (RBR) (25.84 bit/min vs. 23.43 bit/min) than RFP, and could elicit particularly increased ERPs. Moreover, RBFP obtained a high degree of preference in the subjective assessment. In conclusion, the proposed paradigm combined the advantages of both types of stimuli to obtain a significant improvement in BCI performance, which was meaningful for future BCI studies.