Corticomuscular Synchronization Analysis During Stroke Rehabilitation

The research on corticomuscular synchronization has become a hot area in motorneuroscience. The interaction between the motor cortex and the muscles is consideredessential in the process of movement control. This interaction is embodied in thesynchronization phenomenon between electroencephalogram signals (EEG) andelectromyography signal (EMG). Corticomuscular synchronization can reflect thefunctional coupling between the motor cortex and the muscles, and offer tools forunderstanding the underlying mechanisms that produce motor control and motordysfunction. This work provides a way to investigate the collaborative process in neuralnetwork and a new method to evaluate the functional state during neural rehabilitation.Firstly, this article introduced the generation and characteristics of EEG and EMG,and described the synchronized activities in neural system. Consulting with the presentresearch on synchronization analysis method, this work compared the performance andapplication of different algorithms and determined the research orientation, which proceedfrom the consistency synchronization level and generalized synchronization level, on thebasis of coherence analysis and information transfer analysis to study the corticomuscularsynchronization during stroke rehabilitation.Secondly, to address the non-stationary characteristics of neural signals, waveletcoherence analysis was introduced to estimate the time varying correlation between theEEG and EMG. The simulation and measured data analysis proved that wavelet coherenceanalysis can effectively describe the time varying characteristics of corticomuscularcoherence.Thirdly, given the nonlinear characteristics of corticomuscular coupling, this workintroduced informatics tools. Information transfer index (ITI) was defined based oninformation entropy, and due to the different functional role of specific frequency band inneural signals, the ITI was calculated in different frequency bands(Beta and Gamma band)decomposed by wavelet method. The simulation analysis shows that ITI can reliablydescribe information transfer between different coupling models. Finally, experimental study was conducted.6stroke patients and6healthy subjectswere enrolled in our experiments. All subjects were asked to exert a target force. ScalpEEG and sEMG were simultaneously recorded during the task. Coherence method andinformation transfer index was used to analysis the corticomuscular synchronization. Thesynchronization significance level was verified using surrogate data. And the correlationbetween corticomuscular synchronization and different factors was analyzed using statisticmethod. The result show motor function relevance of the synchronization, which providenew methods to evaluate motor function during neural rehabilitation.