Corticomuscular Coherence Based on Wavelet Transform During Imagination, Execution and Stimulation Tasks

The correlation between brain and muscles is an important aspect in the research of motor control. In our study, corticomuscular coherence is used as a method of estimating this correlation. EEG is known to reflect the brain activity, and it is often used to detect or treat some certain illness. EMG, which is related to muscle contractions, is widely used in rehabilitation and ergonomics. These two signals are recorded synchronously to calculate the corticomuscular coherence. Three movement modes, including imagination, stimulation and execution, are explored to reveal the inner mechanism of human's behaviours. Subjects were asked to imagine or execute the flexion of the middle finger, and the electrical stimulation was used for subjects to induce stimulated finger flexion. Wavelet- based coherence is first proposed in our study to observe the coherence in time-frequency domain. During the execution task, 5 of 10 subjects showed significant coherence; 8 of 10 subjects showed strong coherence in the stimulation task. According to the statistical analysis, the coherence during stimulation is stronger than that during imagination and execution. In addition, the significant coherence appeared in different time for different tasks. This shows that stronger relationship between the brain and muscles exists in actual movements. However, how the control paths for execution and stimulation tasks are different still remains unclear according to our result.