Edge-Based Serious Game for Home Rehabilitation
Published:
Duration: 2025
Role: Co-developer; responsible for sensor integration, ML model design, and on-device deployment.
Collaborators: Rehabilitation specialists from UMCG; supervised by Prof. dr. Elisabeth Wilhelm.
Overview
Co-developed an Arduino-Unity-based rehabilitation game for real-time motion and gesture recognition on resource-constrained edge devices. The system is designed for post-stroke and musculoskeletal rehabilitation, addressing the need for accessible home-based exercise tools that do not require clinical supervision.
The game supports customizable rehabilitation targets, on-device model inference, and gamified exercise routines, enabling patients to perform guided rehabilitation exercises at home while receiving real-time feedback on their movement quality.
Demo
Deliverables
- Working prototype demonstrated at the ISCOMS 2025 (International Student Congress of (bio)Medical Sciences) workshop.