EMG Signal Enhancement from ECG Artifact Contamination for Assessment of Swallowing Function

Swallowing disorder can be caused by the inability of muscle groups related in swallowing process to conduct muscle coordination. Swallowing disorder which generally caused by stroke is named dysphagia. Evaluation and diagnosis of swallowing function in patients with the risk of dysphagia is an initial step to be able to treat dysphagia effectively. The surface electromyography (s-EMG) signal measurement had been approved as one of the methods in evaluating muscle activity, which also could be used as an evaluation and diagnosis method of dysphagia. However, the acquisition of EMG signal was generally disturbed by noise from electrocardiography (ECG) artifact. The proposed method, Wavelet-ICA (WICA), was applied in the acquired raw EMG signal to remove the ECG artifacts. WICA was considered based on its reliability in low computational cost, automated decomposition, and high SNR results. Then, the denoised EMG signal was extracted to the signal envelope to determine the muscle coordination parameters. Muscle coordination parameters were observed in s-EMG pattern, which obtained from root mean square value of s-EMG signal envelope. The s-EMG were acquired from bilateral suprahyoid muscle groups. Muscle coordination was evaluated during the subject performed three types of voluntary swallowing tasks with three trials of repetition. According to the results, the higher the cross-correlation coefficient value, the better the muscle coordination in the muscles involved.