Title : Surface EMG profiling and weight influence: Optimizing parkinson’s disease severity assessment with GCN-SVM
Abstract:
This study analyzes variations in surface electromyography (sEMG) signals under different loading conditions to assess their influence on muscle tremors associated with Parkinson’s disease (PD). By investigating amplitude and frequency fluctuations, we seek to enhance understanding of how external loads impact tremor characteristics, ultimately contributing to the development of standardized diagnostic and monitoring criteria. We conduct controlled sEMG measurements with varying loads, systematically assessing tremor behavior in relation to different anthropometric factors and relative maximal strength (RM). Our aim is to refine the understanding of neuromuscular impairments in PD and support the establishment of personalized intervention strategies.
Two opposing hypotheses are being tested: (1) load mass directly affects the measurement quality, necessitating a detailed classification of parameters such as age, gender, body weight, muscle mass, BMI, and physical activity level to establish quantification rules for load adjustment; (2) load mass has minimal impact on measurement outcomes, serving primarily to evaluate biceps engagement during elbow joint movement. In the latter case, a default 2 kg load—previously employed and observed to yield differential tremor amplitudes across disease progression stages—could remain standard.
Unlike the currently used Unified Parkinson’s Disease Rating Scale (UPDRS), which is subject to inter-rater variability and cognitive decline-related inaccuracies, sEMG provides reproducible, quantifiable data. Statistical modeling and machine learning approaches are employed to analyze collected data, identifying potential correlations between tremor severity, load variation, and patient-specific factors. In turn, advanced signal processing techniques will be applied to extract key tremor characteristics, allowing for a comprehensive evaluation of muscle activity patterns. This framework offers a foundation for refining diagnostic protocols and establishing new standards in PD tremor evaluation and disease progression monitoring.
