基于延时等概率符号化传递熵分析的脑肌耦合双向神经信息传递规律研究

The Research of Bidirectional Neural Information Transmission of EEG-EMG Coupling Based on Delay Equal Probability-Symbolized Transfer Entropy

为了有效揭示脑肌双向神经传递机制,解决传统分析方法存在的计算复杂度高、动态特征提取能力差等问题,面向手部运动过程脑肌电耦合特征提取任务,提出了延时等概率符号化传递熵脑肌电耦合分析方法,计算了脑肌电神经信息传递时延规律,进行了运动执行过程激活功能区的关联分析,探索了脑肌耦合强度的时序变化规律。通过在线实验表明:人体左右手的脑肌信息传递具有不对称性,且该传递时延约为20~35 ms;运动执行任务中从脑电到肌电(EEG→EMG)过程比肌电到脑电(EMG→EEG)具有更强的传递熵值。同时,在不同耦合方向下,主动运动任务的脑肌电耦合强度显著高于静息状态。研究不仅对现有的脑肌电耦合分析方法进行了改进,提出了延时等概率符号化传递熵分析方法,同时通过在线实验分析,揭示了手部运动任务下脑肌耦合双向神经信息传递规律,为新的康复方案和康复评价方法提供了理论依据,为神经接口技术的发展提供有力的算法支撑。

The coupling information of EEG and EMG can effectively reveal the neural transmission mechanism and functional interaction information between the cortex and muscles in the process of neuromotor control,and also provide an important theoretical basis for neural interface technology.In order to effectively reveal the mechanism of bidirectional neurotransmission between brain and muscle and solve the problems of high computational complexity and poor dynamic feature extraction ability of traditional analysis methods,this paper proposed a coupling analysis method of EEG and EMG based on time-delay equal probability symbolized transfer entropy for the task of EEG-EMG coupling feature extraction in the process of hand movement.The delay law of neural information transmission between EEG and EMG was calculated,and the association analysis of the activation functional area in the process of movement execution was carried out,the timing variation of the strength of the EEG-EMG coupling was explored.Online experiments showed that the EEG-EMG information transmission of the left and right hands is asymmetrical with the 20—35 ms transmission delay.In movement tasks,the process from EEG to EMG has a stronger transfer entropy than that from EMG to EEG.At the same time,the strength of EEG-EMG coupling of active movement tasks was significantly higher than that in the resting state under different coupling directions.This study optimized the existing EEG-EMG coupling analysis method,and proposed the time-delay equal probability symbolized transfer entropy analysis.Meanwhile,it revealed the bidirectional neural information transmission law of EEG-EMG coupling under hand movement tasks by online experimental analysis,which provided a theoretical basis for new rehabilitation strategy and rehabilitation evaluation methods that support the development of neural interface technology.