39th Annual Student Research Conference
Designing Interactive and Engaging ExerGames to Promote Rehabilitation for Stroke Patients
Enyetullah Rahimullah
Dr. Kafi M. Rahman, Faculty Mentor
This research presents a deterministic, vision-based exergame for post-stroke upper-limb rehabilitation that combines engaging play with reproducible assessment. The goal is to provide an accessible, low-cost system for high-repetition, task-specific training while generating meaningful performance metrics across sessions and against healthy benchmarks. Built with Python, OpenCV, and MediaPipe, the system uses markerless finger tracking with One Euro filtering and rolling median smoothing for stable, responsive control. The game includes a reflex-based task and a guided path-tracing task targeting trajectory control, endurance, and motor accuracy. Deterministic gameplay is achieved through seeded level design and fixed geometric paths, enabling repeatable testing conditions. Performance is evaluated using quantitative criteria, including progress rate and path accuracy, while telemetry data are logged to CSV for analysis. Preliminary healthy-user testing showed technical robustness, low-latency interaction, and sensitivity to difficulty differences, supporting the system’s potential as a measurable rehabilitation platform and a foundation for future clinical validation.
Keywords: Post-stroke rehabilitation, Upper-limb recovery, Vision-based exergame, Deterministic gameplay, Markerless hand tracking, Motor accuracy assessment, Reproducible performance benchmarking, Task-specific training
Topic(s):Computer Science
Presentation Type: Oral Presentation
Session: -1
Location: MG 2007
Time: 10:00