Smart Rehab: AI-Powered Markerless Mobile Rehabilitation for Neuro-Musculoskeletal Recovery

Topic Description

Rehabilitation services worldwide face increasing strain, leaving many patients without regular access to physiotherapy support. Smart Rehab aims to address this by developing an intelligent, AI-enhanced mobile rehabilitation system that enables patients to complete clinically-validated exercises at home — with automated assessment, real-time feedback, and personalised progression. Specifically, this project will:

• Design and build a mobile application that integrates data from the Open University motion-capture lab to monitor user movement quality during rehab exercises.
• Develop algorithms and interaction techniques to assess movement, detect deviations or errors, and provide corrective feedback or recommendations.
• Investigate how to present feedback in an accessible, user-friendly way — for example via visual overlays, simplified metrics, or gamified cues — to sustain motivation and adherence.
• Evaluate the system with real users (patients or volunteers) to assess usability, effectiveness of feedback, and impact on rehabilitation outcomes.

The overall goal is to create a research-driven product) that could significantly improve accessibility, consistency, and effectiveness of physical rehabilitation — reducing burden on healthcare services and enhancing patient recovery outside traditional clinics.

Skills Required

Applicants should have experience or strong interest in:

Essential

• Mobile App Development (Android or iOS or Flutter/React-Native)
• Machine Learning / Deep Learning (PyTorch, TensorFlow)
• Computer Vision (pose estimation, skeleton tracking, image processing)
• Programming (Dart, Java/Kotlin, or Swift)
• Understanding of research methods and experimental evaluation

Desirable

• Interest in physiotherapy, biomechanics, or rehabilitation science
• Experience with motion-capture systems (Vicon, Kinect, OpenPose, MediaPipe)
• Data collection, annotation, and user-study design
• Explainable AI or human-AI interaction

Self-motivated applicants with a passion for AI-driven health technologies are strongly encouraged to apply.

Suggested readings

B. C., I.A., & J.M. (2025). Smartphone-Based Markerless Motion Capture for Accessible Rehabilitation: A Computer Vision Study. Available at - https://www.mdpi.com/1424-8220/25/17/5428 

Ismail-Fawaz, A., Devanne, M., Berretti, S., Weber, J., & Forestier, G. (2025). Deep Learning for Skeleton Based Human Motion Rehabilitation Assessment: A Benchmark. Available at - https://arxiv.org/abs/2507.21018

Park, C., & Lee, B.-C. (2024). A Systematic Review of the Effects of Interactive Telerehabilitation with Remote Monitoring and Guidance on Balance and Gait Performance in Older Adults and Individuals with Neurological Conditions. Bioengineering, 11(5), 460. Available at - https://www.mdpi.com/2306-5354/11/5/460

Multimedia Systems (2022). A review of computer vision-based approaches for physical rehabilitation and assessment. Multimedia Systems, 28, 209–239. Available at - https://link.springer.com/article/10.1007/s00530-021-00815-4

Liao, Y., Vakanski, A., Xian, M., Paul, D., & Baker, R. (2020). A Review of Computational Approaches for Evaluation of Rehabilitation Exercises. Available at - https://arxiv.org/abs/2003.08767