Enhanced EEG-EMG coherence analysis based on hand movements

Electroencephalogram (EEG)-electromyogram (EMG) coherence analysis is an effective method for examining the functional connection between brain and muscles. An improved coherence approach is proposed in this study to enhance the estimation of EEG-EMG coherence. First, we sampled the synchronous EEG signal based on the burst points of the EMG signal. Then, a moving average of the sampled EEG by using a window function is performed before the EEG is sampled again on the basis of the EMG burst points. The EEG signals are reassembled to effectively reflect the muscle motions. Finally, the estimation of the EEG-EMG coherence is computed by using magnitude square coherence (MSC) and wavelet coherence. The coherence characteristics of the different autonomous movements in the beta-band and gamma-band are analyzed to verify the reliability of the method. Results show that our proposed method can remarkably enhance EEG-EMG coherence estimation regardless of using either MSC or wavelet coherence. The results of coherence analysis not only can correctly reflect the coupling relationship between the cortex and the muscles but can also distinguish the EEG-EMG coherences of the different autonomous movements. (C) 2019 Elsevier Ltd. All rights reserved.