Muscle Compensation Analysis During Motion Based on Muscle Functional Network

How to effectively detect muscle compensation caused by exercise fatigue has attracted increasing attention during the last decade, with most previous work being focused on the research of compensatory movement strategy, which indirectly reflecting muscle compensation through the analysis of kinematic data. However, relatively few investigations can be found in the literature for using muscle information to analyze muscle compensation directly. This paper focuses on the muscle compensation analysis based on a 8-channel surface electromyography (sEMG). We investigate the application of complex networks theory in kinematic analysis. Motivated by the good performance of complex networks, we further propose designing the muscle functional network based on sEMG to analyze muscle compensation during exercise, and extract the network features that can reflect the changes of muscle compensation are extracted from the network. Real results from twelve subjects show that muscle functional network can effectively reflect the changes of muscle relationship caused by exercise-induced fatigue. The feature extracted from the muscle network has a high correlation with the feature of muscle fatigue (
$r_{\text{ACC-MDF}}:0.7157$
,
$r_{\text{AGE-RMS}}:0.7877$
), and can be used as a reliable index to analyze muscle compensation changes.