咬合动作相关肌电对稳态视觉诱发电位的典型频率影响

Effects of jaw clench actions on steady-state visual evoked potential at certain typical frequencies

目的基于稳态视觉诱发电位(steady-state visual evoked potential,SSVEP)和肌电(electromyography,EMG)的组合是广泛使用的混合BCI模式。而对于那些只能控制面部肌肉的使用者,咬合动作相关面部肌电通常与SSVEP结合使用。本研究探讨了下颌咬合相关肌电对枕部电极采集到的SSVEP的干扰情况,进而寻找即使在咬合动作下也可同时进行SSVEP识别的刺激频率。方法根据咬合类型,实验分为3个模式组(无咬合、短咬和长咬合)。在每组模式中,受试者同时注视3个闪烁在6.2 Hz、9.8 Hz和14.6 Hz视觉刺激目标。收集枕区4个位点的SSVEP后观察了在3种咬合模式下,各个闪烁刺激的SSVEP响应频谱,并利用典型相关分析方法识别了SSVEP信号,最后统计了准确率。结果当刺激频率低于20 Hz时,即使有以上2种咬合动作,仍然可以避免其对SSVEP的干扰。根据这些信号的频域特征依然可以识别SSVEP。另外,在咬合动作下进行稳态视觉刺激时,SSVEP的识别率仍然很高(无咬合动作:100.0%,短咬:94.7%,长时间咬合:100.0%)。结论通过合理的频率选择和信号处理,即便下颌咬合动作和SSVEP刺激同时发生时,仍可将咬合动作对稳态视觉诱发电位的影响降低,而且达到较高的识别准确率。

Objective Combinations based on steady-state visual evoked potential(SSVEP)and electromyography(EMG)are the widely used hybrid BCIs.For users who are suffering from severe motor impairments and can only control muscles above their necks,the EMG of jaw clench is commonly used together with SSVEP.This article explores the interference with SSVEP from occipital electrodes by the jaw clench-related EMG so that SSVEP with specific frequency can be identified even during occlusal movements.Methods The experiment was divided into three sets based on the jaw clench patterns(no clenches,chew,and long clench).In each set,the subjects used the same visual stimuli,which were realized by the three flashing targets at different frequencies(6.2 Hz,9.8 Hz,and 14.6 Hz).After collecting the SSVEP at 4 sites in the occipital region,the SSVEP response spectrum of each stimulus was observed under the three jaw clench patterns.Then,the SSVEP signal was identified by the canonical correlation analysis method for accuracy statistics.Results Spectrum responses showed that the interference of the jaw clench EMG on SSVEP could be avoided when the stimulation frequency was lower than 20 Hz.SSVEP could be identified based on the frequency domain characteristics of these signals.During steady-state visual stimulation with jaw clenches,the recognition rate of SSVEP was still high(no clenches:100.0%,chew:94.7%,and long clench:100.0%).Conclusions Through proper frequency selecting and signal processing,the influence of the jaw clench movement on the SSVEP can be reduced and a high recognition accuracy can be achieved,even the jaw clench actions and the SSVEP stimulation occur simultaneously.