Discrete Wavelet Transforms Can Detect Shoulder Muscle Fatigue during Light-duty Assembly Tasks

Shoulder musculoskeletal disorders are among the most prevalent work-related musculoskeletal disorders and constitute an important health problem. One of the occupations that are most affected is workers on assembly lines (Leclerc et al., 2004), as these jobs requires repetitive movement, which, as a risk factor, has been found positively associated with shoulder disorders. Repetitive movement is likely to result in muscle fatigue, and many hypothesize that shoulder muscle fatigue is a precursor of shoulder complaints (Takala, 2002). Short time Fourier transforms (STFT) is commonly used to analyze the changes in the frequency domain of surface electromyography (EMG) during muscle fatigue. Muscle fatigue can be demonstrated by a decreased median power frequency. However, STFT is only suitable for analyzing the signals with the same scale. During dynamic movement, the scale of EMG signals is changing so that using STFT to analyze EMG data could produce inaccurate results. Recently, discrete wavelet transform (DWT) has been adopted to analyze the EMG signals for dynamic conditions. The purpose of this study is to investigate whether DWT can be used for detecting shoulder muscle fatigue during light-duty assembly tasks.