基于异步并行诱发策略的混合范式脑-机接口技术

A Technical Study of Hybrid Brain-Computer Interface Based on Asynchronous Parallel Evoked Strategy

稳态视觉诱发电位(SSVEP)与事件相关电位中P300成分相结合的混合范式脑-机接口(SP-BCI)系统可同时诱发两种特征脑电信号并综合前者的高信噪比和异步兼容特点及后者的大指令集优势,具有提高系统信息传输速率的潜在能力,但现有脑电诱发范式未能充分发挥上述特长.本文提出一种SSVEP按各自频率异步诱发和阻断(SSVEP-B)且与P300并行诱发的新策略,并融合SSVEP-B与P300特征信息进行脑电分类识别.经10名健康年轻被试离线测试实验结果表明,被试总体平均分类正确率为84.5!,系统最高理论信息传输速率为89.5 bit/min,表明新型诱发策略有助于提高BCI信息识别正确率和信息传输速率,有关研究思路与技术可供混合范式脑-机接口系统设计与推广应用参考.

A hybrid SSVEP-P300 brain-computer interface （SP-BCI） system combining steady-state visual evoked potential （SSVEP） and P300 component of event-related potential （ERP） could induce both signals at the same time, and take advantage of high signal-to-noise ratio and asynchronous compatibility of SSVEP and the ability to present a large number of commands of P300. It also has the potential to improve information transfer rate （ITR） of the system, but existing evoked paradigms could not give full play to the aforementioned characteristics. This paper proposed a new strategy to make SSVEP evoked and blocked （SSVEP-B） asynchronously according to the respective frequency, to evoke P300 at the same time, and to combine SSVEP-B with P300 to make classification as well. Ten healthy subjects participated in the study. The results of the offline tests show that the system could reach the average accuracy of 84.5% and the highest theoretical information transfer rate of 89.5 bit/min. The results prove that the new strategy is conductive to improve accuracy and information transfer rate of the BCI system, and that related research ideas and technologies can be used as reference to design and generalize a hybrid BCI system.