EGE-EMG Signal Processing And Coherence Analysis For The Neuromuscular Activity Patterns With Three Motion Modes Under Two Time Intervals

The human neuromuscular activity was always accompanied by the release of corresponding electrophysiology signal. In recent years, the cooperative analysis of EEG and sEMG has become a hot research topic in the filed of sports medicine and rehabilitation engineering, etc. The typical patterns of neuromuscular activity include voluntary,stimulation and imaginary motion modes. Because of the differences of brain-control-muscle mechanism, their evoked EEG and EMG also differ in signal pattern. The relationship between sport and brain would be further understood through the comparative analysis of inner link rules between EEG and EMG under difference motion modes.Aimed at one neuromuscular activity example of finger flexion, this thesis designed three kinds of modes including voluntary, median nerve stimulation and imaginary motion and, according the action time intervals, set two experiment programs including the long and short time intervals with 10s and 2s of the time intervals between the adjacent movements, respectively. The EEG from C3/C4 channel and the sEMG from flexor digitorum superficialis were collected simultaneously. After the signal preprocessing including threshold detection of wave peaks between two points, the Morlet wavlet-based time-frequency analysis method were adopted to study the independent variation mechanisms between EEG and EMG under different modes. In this research, partial directional coherent (PDC) algorithm, the extreme points of improved cross-correlation coefficient and time-frequency feature analysis were used to compare and analyze the coherences between the EEG and EMG under different motion modes and time intervals.Research results in this thesis showed that for voluntary motion mode, the contralateral EEG played an obvious domination role to EMG and there was a phenomenon of information interaction between the contralateral and ipsilateral EEG; for stimulation motion mode, EMG was the source of both contralateral and the ipsilateral EEG and showed higher influence on contralateral EEG than ipsilateral EEG; for imaginary motion mode, there was no obvious causality between EEG and EMG, but there still definitely existed the information interaction between the contralateral and ipsilateral EEG. With the same time interval, the causality of the stimulation motion was more significant than voluntary motion while the imaginary motion showed the lowest causality. There was a phenomenon that extreme points of EEG-EMG coherence concentrate inβ-band both in the long and short time interval movement, and the latter is significantly more than the former.In this thesis, through joint analysis of EEG and EMG signals in different time intervals and movement modes, it was found that there was high causality under the short time interval and stimulation mode, which may be relevant to the EEG information overlapping under short time interval movement and more excited brain neutrons and higher dominantion for limbs movement. The research conclusion will contribute to a better understanding of motor and coordination abilities of human neuromuscular system and may provides the references and helps for the clinical rehabilitation treatment protocol in the future.