Uplift-Mobility analysis for post-stroke recovery with exercise tracking and virtual-assisted therapy

ABSTRACT: Stroke remains a leading cause of long-term disability worldwide, significantly impairing individuals mobility and quality of life. Traditional post-stroke rehabilitation methods, while effective, often face challenges related to consistency, patient adherence, and accessibility. This study investigates the impact of modern technological interventions, such as exercise tracking and virtual-assisted therapy, on enhancing post-stroke mobility recovery. The primary objective was to develop the Uplift Mobility mobile application, which offers real-time data collection, personalized exercise programs, and immediate feedback. Utilizing a sprint (scrum) methodology, key features of the application include facial recognition for secure user identification, mobility analysis using MediaPipe for optimizing exercise routines, and an emergency alert system employing audio notifications. The study successfully achieved its objectives, with the facial recognition system implemented using Dlib achieving 95.65% accuracy, and the pose estimation module using MediaPipe reaching 88.25% accuracy. Despite challenges such as limited tool access and connectivity issues, the Uplift Mobility application demonstrated significant potential in supporting post-stroke rehabilitation by enhancing exercise adherence and improving patient outcomes. Future iterations, informed by user feedback, aim to further optimize the application’s functionality and user experience

Filipiniana