Muscle fatigue detection in upper limbs during the use of the computer mouse using discrete wavelet transform: A pilot study. (July 2022)

• Record Type: Journal Article
• Authors: Mota-Carmona, Juan R.; Pérez-Escamirosa, Fernando; Minor-Martínez, Arturo; Rodríguez-Reyna, Reynaldo M.
• Is Part Of: Biomedical signal processing and control. Volume 76(2022)
• Publication Date: 2022-07
• Subjects: Discrete wavelet transform -- Fatigue -- Shoulder -- Forearm -- Repetitive strain injuries -- Computer mouse; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2022.103711
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: Muscle fatigue detection when using the computer mouse was proposed by using DWT.. MNF and MDF extracted features detected better muscle fatigue than power. Coif5 and Db6 wavelets showed the best performance to detect muscle fatigue. Optimum fatigue detection was found in the range of 90 Hz – 130 Hz. Abstract: Background: Due to the current global situation of the pandemic, computer use has increased and has become essential for working and studying. Hence, detection of muscle fatigue associated with computer mouse use is essential to prevent musculoskeletal disorders, since it is considered a precursor of some musculoskeletal injuries. Objective: This study aims to detect muscle fatigue in the shoulder and forearm caused by repetitive and continuous strain associated with the computer mouse, using the discrete wavelet transform (DWT) to prevent musculoskeletal disorders. Methods: Ten participants performed a one-direction tapping task on a computer with four difficulty levels while the shoulder and forearm signals were recorded. We used twelve wavelet functions for DWT analysis. Then, power, MNF, and MDF features were extracted from the wavelet coefficients to detect muscle fatigue. Results: Spectral changes using MNF and MDF of wavelet coefficients of Coif5 and Db6 functions showed significant shiftings towards low frequencies with magnitudes of 20.2 Hz and 26.5 Hz for the shoulder, and 18.9 Hz and 25.6 Hz for the forearm in the fourth decomposition level when the task difficulty increased, indicating muscle fatigue. Conclusion: This study demonstrated that high-precision computer mouse tasks may cause muscle fatigue, and it can be effectively detected by extracting spectral features of the EMG signal to help clinicians and physiotherapists to prevent severe musculoskeletal disorders. In future work, we will optimize the feature extraction method.