Construction and Analysis of Deep Functional Corticomuscular Coupling Effect

Based on electroencephalogram (EEG) and electromyogram (EMG) signals, the function coupling between the cerebral cortex and muscles has been widely studied to evaluate the motor function and reveal various motor control and pathological mechanisms in healthy individuals or patients with movement disorders. However, the effect of the different signal sources on the functional corticomuscular coupling remains unclear. In this study, four different signal source combinations were constructed by EEG and high-density surface EMG (HD-sEMG) signals as well as their reconstructed source signals to analyze the corticomuscular coupling during isometric index finger contraction tasks at different levels of maximum voluntary contraction. A nonparametric coupling model was used to study the effect of deep source signals on changing the coherence magnitude indicators of corticomuscular coupling related to hand movements. The results showed that the reconstruction of EEG and HD-sEMG signals significantly improved the coherence peak and coherence strength under low-level force. However, as the force level increased, only the reconstructed brain source signals demonstrated a more substantial effect. In addition, the coherence peak was positively correlated with the finger force level. This study demonstrated the importance and positive impact of the reconstructed EEG and EMG signals for estimating corticomuscular coherence.