Coupling Analysis of EEG and EMG Signals Based on Transfer Entropy after Consistent Empirical Fourier Decomposition

In brain motor control, the multi-scale functional cortical muscle coupling (FCMC) characteristics between the electroencephalography (EEG) and electromyography (EMG) signals can reflect the functional connection between cortex and muscle. To analyze FCMC accurately, this paper first performs consistent empirical Fourier decomposition on EEG and EMG signals. The decomposition boundaries are set according to EEG functional frequency bands, and then the zero-phase filter sets are constructed, in order to obtain the consistent functional frequency bands for the decomposition. After the decomposition, the FCMC is analyzed by using transfer entropy. Two motor experiments, knee-jerk reflex and inhibited knee-jerk reflex are performed to test. Experiment results show that the most significant for coupling intensity is in the β-band (14–35 Hz) to all subjects. The coupling intensity in EMG→ EEG direction is higher than that in the EEG→EMG direction in the knee-jerk reflex experiment. While the opposite result is obtained in the inhibited knee-jerk reflex experiment. The consistent empirical Fourier decomposition transfer entropy proposed in this paper is not only an accurate coupling analysis method, but also can solve the frequency band inconsistency problem in the multi-frequency band coupling analysis of two signals and make the coupling analysis more accurate.