Functional Corticomuscular Coupling Based on Bivariate Empirical Mode Decomposition - Multiscale Transfer Entropy

Abstract

The functional corticomuscular coupling (FCMC) is a physiological phenomenon to reflect the multilayered characteristics of the information interaction between electroencephalogram (EEG) and electromyographic (EMG) signals. The multilayered characteristics such as local frequency band, complex and multiscale between the brain and muscles are of great significance to understand the cooperative function of the motor-sensory neural network. Though the multiscale transfer entropy (MSTE) method can effectively describe the multiscale and complex characteristics of the coupling signals to some extent, it fails to describe the FCMC on the local frequency band. Therefore, in this study, we combined the bivariate empirical mode decomposition (BEMD) with the MSTE to construct a new model, named bivariate empirical mode decomposition-multiscale transfer entropy (BMTE), to quantify the synchronous coupling between EEG and EMG signals on the local frequency band at different scales. The results show that the FCMC is significant in both EEG→EMG and EMG → EEG directions at betal and beta2 bands during steady-state grip task. Meanwhile, the maximum coupling strength value at beta2 band on different scales alomost occur on the high scales (9–16 scales), and the significant value at betal band was on the lower time scale. Additionally, the coupling strength at gamma band in EEG→ EMG direction is also significant in the higher scale. The results show that the BMTE method can quantitatively describe the local frequency band and multiscale characteristics between the motor cortex and the contralateral muscle in motor control system. This study extends the relative researches on the FCMC.