Electromyography Parameters to Discriminate Hand Osteoarthritis and Infer Their Functional Impact

Abstract

Surface-electromyography (sEMG) allows investigators to detect differences in muscle activation due to hand pathologies. However, its use as a functional indicator and the challenges related to the required normalization have not been fully addressed. This study aimed to use forearm muscle sEMG signals to distinguish between healthy individuals and patients with hand osteoarthritis (HOA). sEMG data were collected from seven sensors on the forearms of twenty-one healthy women and twenty women with HOA during the Sollerman test. Amplitude-based parameters (median and range) were normalized using three methods: maximum signals during Sollerman tasks (MAX), during maximum voluntary contraction tasks (MVC), and during maximum effort grasping (GRASP). Waveform parameters (new-zero-crossing and enhanced-wavelength) were also considered. MVC and GRASP resulted in higher values in patients. Discriminant analysis showed the worst success rates in predicting HOA for amplitude-based parameters, requiring extra tasks for normalization (MVC or GRASP), while when using both amplitude (MAX) and waveform parameters and only Sollerman tasks, the success rate reached 90.2% Results show the importance of normalization methods, highlight the potential of waveform parameters as reliable pathology indicators, and suggest sEMG as a diagnostic tool. Additionally, the comparison of sEMG parameters allows the functional impact of suffering from HOA to be inferred.