A Novel Stacked Regression Method for Telemonitoring Parkinson’s Disease Outcome Prediction Using Biomedical Voice Features

Parkinson’s Disease (PD) is a complex neurodegenerative disorder characterized by both motor and non-motor symptoms. The development of traditional diagnostic approaches for detecting PD has limitations in terms of accuracy and accessibility, particularly in regard to continuous monitoring. This paper make an effort by proposing a stacked regression model combining Random Forest (RF) and Extreme Gradient Boosting (XGBoost) with an effort of improving predictive accuracy for Unified Parkinson’s Disease Rating Scale (UPDRS) using biomedical voice data as a non-invasive monitoring method. The stacked model achieved lower MSE and higher accuracy compared to individual models demonstrating superior predictive performance. Thus, the combined model effectively captures nonlinear relationships and reduces overfitting through the balanced bias-variance tradeoff with Mean Squared Error (MSE) score of 25.027 an R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of 0.621 offering a scalable and non-invasive solution for predicting UPDRS scores. Overall, the proposed approach has significant implications for real-time monitoring and proactive patient management in PD with potential applications across other neurodegenerative diseases.