Investigating the Effect of Biofeedback on Tremor Reduction in Parkinson's Patients by Identifying Motor Units Roles with Machine Learning

Parkinson's disease, a neurological condition primarily affecting motor function, often manifests through debilitating tremors, significantly impacting patients' quality of life. This study investigates the potential therapeutic role of biofeedback in managing Parkinson's tremors. Biofeedback was implemented and evaluated as a means to alleviate tremors in affected individuals. The analysis focused on surface electromyography (sEMG) signals in the frequency domain, employing Adaboost algorithms to detect changes in motor unit activity. Results indicated a shift in frequency towards lower frequencies, suggesting an improvement in the sensory-motor relationship. Additionally, increased signal amplitude suggests a modification in motor unit patterns, enhancing both task accuracy and muscle contraction power.