Research On Brain And Muscle Electrophysiological Network Related To Upper Limb Movement

The movement of the upper limb is the lever movement produced by nerve innervated muscles contracting and pulling connected bones,and is the most common way of movement in daily life.It is.There are a large number of people with abnormal upper limb function caused by stroke and other factors.For the upper limb functional rehabilitation of these patients,it is of great social significance to be able to reshape brain motor function and restore voluntary motor control.However,there is currently a lack of systematic research on the information interaction relationship of the central nervous system and the peripheral nervous system related to movement.Hence,this study quantitatively describes the information interaction mechanism of these two different nervous systems under upper limb movements from the network level,and studies the electroencephalogram(EEG)network-related mechanism,electromyography(EMG)network and corresponding recognition model,EEG-EMG static coupling network,EEGEMG dynamic interaction network and EEG-EMG causal driving network.The main work carried out in this dissertation is as follows.1.In the central nervous system,existing studies mainly focus on movement-related EEG activities,but there are few studies on the impact of individual long-term strength movement on the resting state of the brain.Therefore,we explored the correlation between upper extremity maximum grip strength levels and resting-state EEG networks.The study found that in the Beta and Gamma bands,the resting-state EEG network connection(especially the frontoparietal network connection)and its network properties were significantly correlated with the individual’s maximum voluntary contraction(MVC),and the model predicted MVC was significantly positively correlated with the actual measured MVC.This suggests that the resting-state EEG network is closely related to upper limb strength,which will help to indirectly reflect the status of individual muscle strength through the resting brain network.2.In the peripheral nervous system,when the corresponding movement is completed,the interaction between various muscle units is not a simple one-to-one interaction,but there is coordination among multiple muscle masses.By constructing the corresponding EMG network,the interaction mode of information between muscle masses can be effectively revealed.Therefore,aiming at the need for convenient collection of multiconductor signals required by the construction of the EMG network,we designed a new type of epidermal array electrode sleeve to collect EMG signals,constructed an EMG network corresponding to upper limb movements through signal coherence analysis,and developed an upper limb movement recognition algorithm based on the EMG network.Finally,a wearable master-slave robot-driven rehabilitation system based on the new type of epidermal array electrode sleeve and network recognition algorithm was developed and verified online.The research results show that the new electrode can collect EMG signals with high quality,and different movements have different EMG spatial network patterns,and based on this network pattern,movement recognition can be performed robustly and reliably.The online recognition accuracy of the upper limb movement of the robot system can reach 98%,and the real-time control of the master hand to the slave hand is realized.These research conclusions provide new theories and potential implementations in the field of neuroprosthetic control and hemiplegic hand rehabilitation.3.Compared with a single central or peripheral nervous system,there is currently a lack of systematic information coupling research on the relationship between the two.For this reason,we constructed the static coupling network within and between the brain and muscles under different thumb forces through coherence analysis.The study found that,in the internal network of EEG or EMG,high-force activity evokes stronger and more concentrated network patterns than low-force activity.In the EEG-EMG interaction network,high force is also stronger than low force in the coupling relationship between motor non-contralateral brain regions and most muscle groups.In the identification of different forces,the EEG-EMG network features have significantly improved compared with single EEG or EMG network features,and the classification accuracy rate has increased by at least 4%.This reveals that under different hand forces,the information interaction coupling between EEG and EMG is different in connection strength and spatial coupling mode,indicating that there is information interaction between the central nervous system and the peripheral nervous system during movement.4.During movement,the information interaction between the central and peripheral nervous systems is constantly changing,and it is also necessary to understand how the brain and muscles are dynamically coordinated.For this reason,we used an adaptive directed transfer function(ADTF)to construct an EEG-EMG dynamic interaction network under different upper limb movement stages.The study found that the dynamic interaction network between the central nervous systems and the peripheral nervous systems exhibits a time-varying information interaction trend during movements.During the movement preparation stage,the muscles dominate and transmit bottom-up motor information;during the movement execution stage,the brain dominates and issues topdown control commands;during the movement ending stage,the dominance of the brain and muscles tends to balance.The recognition model constructed based on dynamic interactive network indicators can realize the classification of different movement stages of upper limb movement,and the recognition accuracy can reach 74%.This study helps to deepen our understanding of the central-peripheral nerve dynamic coordination mechanism,and also provides a theoretical basis for the monitoring and regulation of neural pathways in movement disorders.5.Motor hemiplegia is caused by the interruption of neural pathway connections between the motor cortex and muscle tissue.The interrupted neural pathways can be rebuilt through different rehabilitation modes of upper limb exoskeleton equipment.Among them,the active mode is that the upper limb movement drives the exoskeleton training,and the passive mode is that the exoskeleton drives the upper limb movement training.However,it is not known which rehabilitation mode is more effective in activating the corticomuscular connection.Therefore,we constructed EEG-EMG causal driving networks in active and passive modes by Granger causality(GC)analysis.The study found that the active mode was significantly stronger than the passive mode in terms of the strength of the internal network connections in brain regions contralateral to movement or in movement-related muscles.In the closed-loop regulation process of information transmission-feedback in the corticomuscular neural pathway,the active mode also exhibits a very significant regulation effect.Based on the network features of EEG,EMG,EEG-EMG,and their fusion features,the classification of active and passive modes can be realized,and the recognition accuracy based on fusion features is the highest.This indicates that the active mode can more effectively regulate the corticomuscular information neural pathway,and various electrophysiological networks analyzes also provide neurofeedback for exoskeleton-based rehabilitation robot training for physicians treating motor hemiplegia.In summary,for the study of brain and muscle electrophysiological networks related to upper limb movements,the research value of this dissertation is to explore the relationship between the resting EEG network and MVC,develop the upper limb movement recognition with EMG network as the core,reveal the difference of EEG-EMG coupling network under different forces,explain the EEG-EMG dynamic coordination mechanism during movement,and discover the regulation effect of exoskeleton on EEGEMG neural pathway.