Research On EEG-EMG Coupling Analysis Method Of Upper Limb Movement

Electroencephalogram(EEG)and electromyogram(EMG)signals are the most important research objects in the field of neuroscience and health assessment.The EEG signals contain the control information of the human cerebral cortex on the movement of the limbs,and the EMG signals contain the response information of the limb muscles to the movement control.The EEG and EMG signals do not act separately,there is a complex coupling relationship between them.Therefore,it is necessary to combine EEG and EMG signals to conduct in-depth research on EEG-EMG coupling,especially the time-frequency domain coupling analysis of EEG and EMG signals can reflect the complex nerve conduction interaction information between the two systems,and has important theoretical value and practical significance for more comprehensive evaluation of the functional status of human neurorehabilitation exercises.This paper first introduces the generation principles and characteristics of EEG and EMG signals,then analyzes the development status of EEG-EMG coupling analysis methods,and then describes the pretreatment methods of synchronized data acquisition and analyzes its principles.In order to improve the shortcomings of the existing EEG-EMG coupling analysis methods,this paper separately analyzes the EEG and EMG signals of healthy people and stroke patients under different grip actions during upper limb exercises in time and frequency domain.The main tasks are as follows:(1)The Empirical Mode Decomposition(EMD),Multivariate Empirical Mode Decomposition(MEMD)and Adaptive Projection Intrinsically Transformed MEMD(APITMEMD)and other methods are studied at the level of method principle.In the next step of exploration and research work,it will lay a theoretical foundation for the EEG-EMG coupling analysis.(2)Aiming at the fact that Transfer Entropy(TE)cannot ignore the instantaneous effects in actual time series analysis,a new Multi-scale compensation TE(Mc TE)analysis method is proposed.Synchronously collect the EEG and EMG signals of healthy subjects and patients in the upper limb movement,use APITMEMD method for multi-scale decomposition,and then calculate the Mc TE values in different frequency bands.The results point out that Mc TE can quantitatively describe the cortical functional coupling strength and information transmission between different coupling directions and multi-scales of EEG-EMG signals,provide a theoretical basis for the rehabilitation evaluation of stroke patients.(3)Aiming at the traditional Partial Directed Coherence(PDC)and generalized Partial Directed Coherence(g PDC)cannot perform linear and nonlinear EEG-EMG coupling analysis at the same time,a time-frequency domain synchronous analysis method based on APITMEMD-r TVg PDC is proposed.In the case of eight healthy subjects with different grip strengths,the EEG and EMG signals were collected simultaneously,and then the method was used to study and analyze the characteristics of cortex-muscle coupling and the direction of information transmission.The results point out that this algorithm can describe the linear and nonlinear interactions between EEG and EEG signals under different grip strengths,and provide effective physiological parameters indicators for the study of motor dysfunction and rehabilitation evaluation.