Digital monitoring of motor function in Parkinson's disease using Markerless motion analysis and exergaming

Motor impairment in Parkinson's disease (PD) significantly compromises functional independence. While continuous rehabilitation is crucial, traditional models face logistical and economic barriers that limit continuity of treatments.

To address this challenge, we developed a novel exergaming platform leveraging Google MediaPipe for markerless, real-time kinematic tracking via a standard webcam, eliminating the need for specialized hardware and delivering engaging, gamified physical exercises designed for domestic settings. This study investigates the feasibility, usability, and preliminary clinical impact of a 10-session gaming protocol in 14 out-of-hospital patients.

The system showed high technical performance and participant engagement, with an overall trial completion rate exceeding 94% and successful progression through game levels. We observed improvements in key functional parameters, establishing a strong correlation between level progression, measured by the novel Normalized Efficiency Index, and the clinical MDS-UPDRS assessments, both for total (ρ = -0.61) and mobility (ρ = -0.66) scores. Furthermore, the system detected performance incongruence related to medication timing and motor fluctuations. In addition, a session-by-session analysis revealed consistently high patient satisfaction, engagement, and system usability scores, alongside low perceived physical fatigue.

These findings underscore the clinical validity and high acceptance of the proposed solution as a training and remote monitoring tool. By providing granular, longitudinal data, this highly accessible solution offers a promising approach to personalized home-based functional training for people with PD.