SURFACE ELECTROMYOGRAPHY QUANTIFICATION FOR EVALUATING MUSCLE COORDINATION IMPROVEMENT ON SWALLOWING DISORDER REHABILITATION SCHEME

• Main Author: Syifalianti Noor, Azizah
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/46923
• Institution: Institut Teknologi Bandung

The current stroke event is transitioning from the cause of death to the cause of disability. It happens since the deaths due to stroke decreases, while the number of disabilities due to stroke increases. The data show that 44-63% of acute stroke sufferers experience swallowing disorders or dysphagia as a consequence of disability from stroke symptoms. Quantitative evaluation of stroke patients with the risk of swallowing disorders is needed to increase the accuracy in determining the diagnosis and further rehabilitation scheme needed by the patient. Conventionally the use of fluoroscopy X-ray videos or video-fluoroscopic swallowing studies (VFSS) is used for diagnosis. Another approach that developed without radiation risk is based on the pattern of muscle coordination involved in the swallowing process as an indication of the severity of symptoms of swallowing disorders in patients. In this study, surface electromyography (sEMG) is used to quantify the coordination of muscles involved in the swallowing process, which measured in the suprahyoid (SUP) and infrahyoid (INF) muscle groups. A total of 30 healthy subjects and 44 post-stroke patients (mean age 39.8 ± 15.5 years) were involved in this study. The experimental scheme instructs the subject to swallow 3 mL of water (normal swallow), 15 mL (effortful swallow), and swallow saliva (dry swallow). Measurement of sEMG signals during the swallowing process in the neck area is often influenced by noise and heart rate artifacts (electrocardiography-ECG). The decomposition of contaminated sEMG signals was performed by the discrete wavelet transform (DWT) method. The analysis of swallowing conditions is determined based on the features of the sEMG signal in healthy subjects and stroke sufferers at risk of swallowing disorders. The sEMG signal features consisting of duration, onset, offset, time to peak, root means square (RMS), cross-correlation, and time-frequency spectrogram are significantly able to distinguish swallowing processes in healthy subjects and post-stroke patients. The results show that the duration of swallowing in healthy subjects has a smaller value compared to patients. The time to peak parameter in healthy subjects shows a faster value compared to patients. The faster duration of swallowing and time to peak value show a better delivery period of food or drink from the mouth into the stomach. The strength of the sEMG (root mean square-RMS) signal indicates that the suprahyoid muscle is the more dominant muscle in the swallowing process. The average RMS value in suprahyoid muscle is higher than the RMS value in infrahyoid muscle. A higher RMS value indicates a more significant strength of muscle contraction. Quantification of muscle coordination on the sEMG signal is represented as the coefficient of cross-correlation and time-frequency spectrogram. The higher value of these parameters, the better muscle coordination occurred. The correlation coefficient and spectrogram area in healthy subjects shows a higher value compared to post-stroke patients. It concludes that healthy subjects show better muscle coordination compared to post-stroke patients. Analysis of dysphagia severity is significanty identified by sEMG features namely RMS, cross-correlation coefficient and time-frequency spectrogram. The severity of dysphagia is classified into healthy, stroke non-dysphagia and stroke with dysphagia. RMS features threshold on SUP muscle of healthy subjects, stroke patient non-dysphagia and stroke patient with dysphagia were obtained at 23.0658 mV, 21.0173 mV and 18.7766 mV (normal swallow) also 24,4743 mV, 17,1983 mV and 22,8421 mV (dry swallow). Cross-correlation coefficient threshold of healthy subjects, stroke patient non-dysphagia and stroke patient with dysphagia were obtained at 0.9073, 0.8068 and 0,8049 (normal swallow) also 0.9234, 0.8120 and 0.7994 (dry swallow). Normalization of spectrogram area on SUP muscle of healthy subjects, stroke patient non-dysphagia and stroke patient with dysphagia were obtained at 1.8604, 1.3040 and 1.1794 (normal swallow) also 2.3240, 1.1288 and 1.1142 (dry swallow). This identification has the potential to be used as a reference for clinicians and doctors to suggest further rehabilitation schemes that will be applied to patients after stroke, make a diagnosis and prognosis of the severity of symptoms of swallowing disorders (dysphagia).