ARMStrokes: A Mobile App for Everyday Stroke Rehabilitation
Jin Guo, Ted Smith, David Messing, Ziying Tang, Sonia Lawson, Jinjuan Heidi Feng · 2015 · ASSETS '15: Proceedings of the 17th International ACM SIGACCESS Conference on Computers & Accessibility · doi:10.1145/2700648.2811337
Summary
This paper presents ARMStrokes, an iPhone application that enables stroke survivors to perform upper extremity rehabilitation exercises anywhere and anytime using only built-in smartphone sensors. Stroke affects over 795,000 Americans annually and often causes long-term disabilities, yet only 31% of survivors complete recommended rehabilitation exercises—largely due to lack of motivation. While gaming systems like Wii have shown rehabilitation benefits, seniors (who comprise most stroke survivors) may lack access to these systems or the physical capability to use them. ARMStrokes uses the iPhone's accelerometer, gyroscope, and orientation sensor to detect arm movements, requiring no additional hardware. The app includes eight therapist-selected exercises targeting upper limb muscles and joints: forearm rotation, elbow flexion, elbow raises (front and side), shoulder flexion, shoulder rotation, shoulder horizontal adduction, and shoulder abduction. Two game metaphors were developed—a monkey collecting bananas and an astronaut exploring space—where the character's movements mirror the user's detected arm motions. The app provides multimodal feedback (visual cues, audio, haptic vibration) that users can customize. A key design feature is personalization: therapists configure daily goals specifying exercise types, sets per day, and duration. Functional calibration allows the system to detect movements from patients with very limited range of motion or slow speed, and can identify when patients inappropriately use their unaffected arm.
Key findings
Preliminary evaluation through four focus groups involving twelve stroke survivors and their caregivers, three therapists, and one physician yielded highly positive feedback. After approximately 10 minutes of training, users could navigate the app independently and complete exercises. All participants expressed enthusiasm about using ARMStrokes daily. Stroke survivors specifically noted that home rehabilitation exercises were "quite boring" and the app would motivate them to complete more exercises. The instant feedback after each exercise was particularly valued. The rehabilitation plan and calibration settings need updating as patients' functionality gradually recovers, enabling the app to grow with the user through their recovery journey. The research team was collaborating with the University of Maryland Rehabilitation and Orthopaedic Institute on a two-month field study with 50 outpatients to determine whether the app positively impacts stroke recovery. Customized solutions (straps, gloves) were being explored to assist with appropriate phone placement based on individual movement capabilities.
Relevance
This research addresses a critical gap between clinical stroke rehabilitation (which has limited duration and frequency) and home-based exercise (which suffers from low compliance). The smartphone-based approach eliminates barriers that prevent seniors from using gaming rehabilitation systems: no additional hardware to purchase, no complex setup, and exercises can be performed while seated. The emphasis on therapist-configurable customization reflects real clinical workflow—therapists assess functional capabilities, prescribe appropriate exercises, and adjust parameters as patients recover. This positions the technology as a complement to professional care rather than a replacement, increasing likelihood of clinical adoption. For accessibility practitioners, ARMStrokes demonstrates how consumer smartphone sensors can enable clinical-grade motion tracking for rehabilitation. The multimodal feedback options (visual, audio, haptic) and ability to disable animations address diverse sensory needs among stroke survivors, who may have visual or cognitive impairments alongside motor deficits. The cheating detection feature (identifying use of the unaffected arm) shows attention to the clinical validity of self-administered therapy.
Tags: stroke rehabilitation · mobile health · exergames · upper extremity · motor impairment · physical therapy · gamification · smartphone sensors · telerehabilitation · occupational therapy