Smartphone-based Rehab Platform for Post-Stroke Upper Limb Training

Thesis Abstract

Every year, a large proportion of stroke survivors experience persistent upper-limb impairments that limit daily functioning, yet access to engaging, intensive, and affordable rehabilitation remains inadequate in many settings. This study addresses the gap by developing and evaluating a smartphone-based rehab platform designed to augment post-stroke upper-limb training through multimodal motor practice, real-time feedback, and remote monitoring. The aim is to determine the platform’s effectiveness in improving motor outcomes, adherence, and user experience, and to explore factors that influence uptake and adherence. Specific objectives include (1) assessing changes in upper-limb motor impairment using the Fugl-Meyer Assessment for the Upper Extremity (FMA-UE) and the Box and Block Test over a 12-week protocol; (2) evaluating functional performance in activities of daily living via the ADL Summary Score; (3) examining adherence rates, session intensity, and dose–response relationships; (4) analyzing user satisfaction, perceived usability, and engagement through mixed-method data; and (5) identifying demographic, neuropsychological, and technology-literacy predictors of outcomes. The study employs a randomized controlled design with a parallel-group format, enrolling 150 community-d dwelling stroke survivors aged 40–85 years within 6–12 months post-onset. Participants are randomized to receive either standard care plus the smartphone-based rehab platform (intervention, n=75) or standard care alone (control, n=75) for 12 weeks. The platform integrates sensor-backed exercises, gamified tasks, contextual biomechanical feedback, and remote clinician monitoring, grounded in Motor Learning Theory and the Biopsychosocial Model to address motor recovery, motivation, and participation. Data collection comprises baseline, mid-intervention (6 weeks), and post-intervention (12 weeks) assessments, including (a) primary outcomes FMA-UE and Box and Block Test; (b) secondary outcomes ADL Summary Score, grip strength, and kinematic metrics captured via smartphone sensors; (c) process measures adherence logs, time-on-task, perceived exertion, and fatigue scales; and (d) qualitative data from semi-structured interviews with a purposive subsample of 20 participants and 8 clinicians. Quantitative analysis uses repeated-measures ANOVA to examine changes over time between groups, mixed-effects linear models to account for heterogeneity, and regression analyses to identify predictors of response and adherence. Mediation analysis explores whether engagement mediates the relationship between platform use and motor outcomes. For qualitative data, thematic analysis follows Braun and Clarke’s approach, triangulated with quantitative findings to elucidate user experiences, barriers, and facilitators. The study also employs a cost-effectiveness assessment from a healthcare payer perspective, estimating incremental cost per additional functional recovery point on the FMA-UE. Expected findings include statistically significant improvements in motor impairment and manual dexterity in the intervention group versus control, higher daily training dose with positive correlations to outcome gains, and favorable usability and acceptance scores with acceptable adherence (mean adherence ?70%). It is anticipated that greater improvements will be observed in participants with higher baseline technology familiarity but that the platform will be usable across a broad age range, given tailored instructional materials. The study contributes to knowledge by providing rigorous evidence on a scalable, ICT-driven rehabilitation solution that integrates motor learning principles, objective sensor-based monitoring, and remote clinician oversight to augment conventional therapy. It advances understanding of dose–response relationships in telerehabilitation for post-stroke upper limb recovery and identifies patient and programmatic factors that optimize engagement and outcomes. The main conclusion is that smartphone-based rehabilitation can produce meaningful motor and functional gains when combined with standard care, with sustained adherence supported by user-centered design and clinician engagement. Recommendations include integration into community-based stroke care pathways, optimization of adaptive difficulty and feedback algorithms, expansion to multilingual content, and further research on long-term maintenance of gains and applicability to individuals with varying cognitive and sensory impairments.