A high-frequency SSVEP-BCI system based on a 360 Hz refresh rate

Ke Liu,Zhaolin Yao,Li Zheng,Qingguo Wei,Weihua Pei,Xiaorong Gao,Yijun Wang
DOI: https://doi.org/10.1088/1741-2552/acf242
IF: 5.043
2023-08-23
Journal of Neural Engineering
Abstract:Objective. Steady-state visual evoked potential (SSVEP) based brain-computer interfaces (BCIs) often struggle to balance user experience and system performance. To address this challenge, this study employed stimuli in the 55 - 62.8 Hz frequency range to implement a 40-target BCI speller that offered both high-performance and user-friendliness. Approach. This study proposed a method that presents stable multi-target stimuli on a monitor with a 360 Hz refresh rate. Real-time generation of stimulus matrix and stimulus rendering was used to ensure stable presentation while reducing the computational load. The 40 targets were encoded using the joint frequency and phase modulation method. Offline and online BCI experiments were conducted on 16 subjects using the task discriminant component analysis algorithm for feature extraction and classification. Main Results. The online BCI system achieved an average accuracy of 88.87±3.05% and an average information transfer rate of 51.83±2.77 bits/min under the low flickering perception condition. Significance. These findings suggest the feasibility and significant practical value of the proposed high-frequency SSVEP BCI system in advancing the visual BCI technology.
engineering, biomedical,neurosciences
What problem does this paper attempt to address?