The existing steady-state visual evoked potential (SSVEP)-based brain-computer interfaces (BCIs) primarily use central visual field flickers with a stimulus frequency of 8-20 Hz, which is prone to exhibit strong flicker perception in users. Considering that, this study aims to develop an SSVEP-based BCI system which is both high-performance and low-flicker-perception by employing high-density electrodes and high-frequency flickers in the peripheral visual field. A custom-made electroencephalogram (EEG) cap with high-density electrodes was used to acquire more EEG data. To alleviate flicker perception, this study combined high-frequency visual stimulation with peripheral visual field stimulation. The proposed system encoded 40 targets using annuli with an angular range in 2.1°-4.1° and high-frequency flickers in the range of 32.00-36.68 Hz. For signal decoding, the task-discriminant component analysis (TDCA) was first applied to the peripheral visual field SSVEP-based BCI system with weak response. Through online experiments, the feasibility of this system was verified. It achieved an average classification accuracy of 83.22 ± 11.95% and an information transfer rate (ITR) of 178.21 ± 43.84 bits/min. Moreover, the role of high-density electrodes to obtain more useful EEG information and thus improving the classification accuracy has been proved. Comparison with existing methods: The online ITR of this system was the highest for current peripheral visual field SSVEP-based BCIs. The proposed system not only provides novel ideas for the design of BCI systems with weak flicker, but also provides reference value for the future application of high-density electrodes in SSVEP-based BCIs. • TDCA algorithm decoded weak SSVEP in peripheral visual field BCIs for the first time. • High-frequency (32-36.68 Hz) and peripheral visual stimulation were used to alleviate flicker perception in SSVEP-BCIs. • The role of high-density electrodes in acquiring more useful EEG data and classification accuracy had been validated. • The 40-target system achieved the highest ITR (178.21 bits/min) in peripheral visual field SSVEP-BCIs.
Pang et al. (Sun,) studied this question.