Early-stage Parkinson's disease classification using biomechanical analysis and machine learning during sit-to-walk task

Early-stage Parkinson’s disease (PD) involves impaired motor control, particularly during complex tasks requiring coordinated postural adjustment and locomotion. The sit-to-walk (STW) task, which integrates standing and gait initiation, presents a substantial fall risk for individuals with PD. However, its utility for early detection and symptom monitoring remains insufficiently investigated. This study aimed to identify STW-related biomechanical biomarkers of early-stage PD and evaluate the performance of a machine learning based classification model. A total of 106 participants were enrolled: 63 individuals with early-stage PD and 43 age-matched healthy controls. Three-dimensional motion capture, force plates, and surface EMG were used while participants performed the STW task at a self-selected speed. From kinematic, kinetic, and neuromuscular data, 200 biomechanical variables were extracted across three task phases. Weighted feature importance and stepwise binary logistic regression identified three key variables: mean center of mass (COM) speed during the entire task; anteroposterior center of pressure (COP) – COM displacement during Phase 2; and forward thoracic range of motion during Phase 2. A random forest classifier using these variables achieved 84.9% accuracy. These biomarkers reflect compensatory movement strategies for maintaining postural stability and may support objective early screening of PD in clinical practice.