Kinerehab: A Kinect-based System for Physical Rehabilitation — A Pilot Study for Young Adults with Motor Disabilities
Jun-Da Huang · 2011 · Proceedings of the 13th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS 2011) · doi:10.1145/2049536.2049627
Summary
This paper presents Kinerehab, a Kinect-based intelligent rehabilitation system designed to assist young adults with motor disabilities in performing physical therapy exercises. Developed in collaboration with physical therapists at the Kaohsiung County Special Education School in Taiwan, the system addresses two interconnected problems: students with physical disabilities (primarily cerebral palsy and muscle atrophy) typically lack motivation to engage in repetitive rehabilitation exercises, and therapists are stretched thin — one therapist typically works with five students but can only actively monitor one at a time. The system uses Microsoft's Kinect motion sensor to detect students' joint positions and automatically determine whether their movements meet rehabilitation standards (e.g., raising both hands a specified distance above the shoulders). It combines several components: automatic movement detection, step-by-step video instruction that pauses and waits for the user to complete each movement before proceeding, voice reminders, a personalised movement menu that therapists can adjust for individual students, and automatic recording of rehabilitation progress. The approach leverages mainstream consumer technology rather than specialised clinical equipment, making it more accessible and cost-effective than traditional virtual reality rehabilitation systems that require wearing body-mounted sensors.
Key findings
A pilot study was conducted with four participants aged 16-17 at the special education school — three with cerebral palsy and one with muscle atrophy. The rehabilitation protocol involved three upper-limb movements (lifting both arms to the front, to the side, and upwards), each repeated six times per cycle for three cycles. The Kinect system achieved over 80% accuracy in detecting movements correctly. The primary source of error was interference from wheelchairs and walkers, which the Kinect misidentified as part of the students' bodies. All four students were highly engaged and wanted to continue using the system after the experiment ended. Both therapists and students reported that the system increased motivation to participate in rehabilitation. The therapist assessed the system favourably, noting it would reduce her workload and improve rehabilitation effectiveness. Student feedback included a request for simultaneous multi-user support to enable peer encouragement during exercises. The therapist suggested incorporating game elements to further enhance engagement.
Relevance
This pilot study demonstrates how consumer motion-sensing technology can be repurposed as accessible, low-cost assistive technology for physical rehabilitation. The findings are relevant to accessibility practitioners in several ways. First, the system illustrates the principle that mainstream technology integration reduces barriers to adoption — students could engage with a familiar, non-stigmatising device rather than clinical equipment. Second, the study highlights an important accessibility challenge: the Kinect's reduced accuracy when users had wheelchairs or walkers shows how motion-sensing systems can fail when they don't account for mobility aids, a design consideration relevant to any gesture-based interface. Third, the work underscores the value of co-design with therapists and end users in creating effective assistive technology. The small sample size and lack of a control group limit the generalisability of results, but the positive engagement outcomes suggest that gamified, technology-assisted rehabilitation merits further investigation.
Tags: physical rehabilitation · motor disability · Kinect · motion sensing · cerebral palsy · assistive technology · gamification