Electromyography signal analysis using time and frequency domain for health screening system task

Abstract

Musculoskeletal disorder (MSDs) is one of the most popular issues of occupational injuries and disabilities. It has a big impact and creates a big problem for industries to be resolved. In MSDs, electromyography (EMG) is one of the methods to be studied in order to detect MSDs problem. This research focuses on the EMG signal analysis by using time domain and frequency domain (Welch Power Spectral Density) method. It gives more information from the signal and it is the most suitable method for classifying the moments in order to identify the behavioural of the signals. Axial rotational reach and upper level reach task from Health Screening Program (HST) is performed using functional range of motion (FROM) by considering left and right biceps brachii muscles to be analysed. There are two parameters chosen for each time and for each frequency domain to be tested, which are mean an absolute value (MAV) and root mean square (RMS) for time domain. Median frequency (MDF) and mean frequency (MNF) are for frequency domain. The results showed that frequency domain analysis is able to give more parameter and information of the signal. Upper level reach acquires more effort to perform the task compared to axial rotational reach for left and right biceps brachii. However, different performances of the signal obtained in classifying the moments from t-test analysis due to p-value. The best performance to classify signal characteristics is the lowest p-value which is 7.369E-05 (MAV), 6.9504E- 05 (RMS), 0.0054 (MDF). However, p-value for 0.0515 is rejected because it is greater than 0.05. It is concluded that the frequency domain is able to give more information of the signal, however for classifications moments, time domain is better compared to the higher accuracy result. This study is very important to give the idea in the future analysis of EMG signal in the aspect of detecting MSDs in human body in health screening task.