Patients undergoing rehabilitation often need repeated follow-up so clinicians can assess movement, spasticity, and recovery over time. For patients who live far from a specialist centre, the assessment itself may be brief while the journey requires many hours, additional expense, and assistance from family members. Requiring every routine review to happen in person can therefore make consistent follow-up difficult, especially when travel is physically demanding.
Many aspects of movement can be recorded with an ordinary camera. A structured video assessment could allow clinicians to review selected movements remotely and decide when an in-person examination is necessary. The challenge is to make those recordings clinically useful: the movement has to be visible, the task has to be performed consistently, and the video has to contain enough information for meaningful comparison across visits. Establishing that foundation comes before attempting automated grading.
Clinicians need a simple way to review routine rehabilitation progress without requiring every patient to travel to hospital for every follow-up. The recording process has to be practical outside a motion laboratory, use equipment that patients or caregivers can reasonably access, and capture movements in a consistent enough manner for comparison over time. Clinicians also need to know when video is insufficient and an in-person examination is still required. At this stage, the priority is therefore not an automated score, but a reliable video-collection process that produces clinically interpretable recordings and can support future model development.
This project is in its initial data-collection and protocol-development phase. We have not trained, validated, or deployed a model, and there are no performance results to report; the work is focused on establishing a video-collection process that captures clinically relevant movements consistently enough to compare across visits, which will then determine whether automated analysis is justified.