CAREER: Enhancing Self-Directed Hand Rehabilitation with AI-Driven Recovery Dynamics Monitoring and Motivation Boosting

Arizona State University, Scottsdale AZ

Investigators

Abstract

A significant number of individuals face challenges with hand movement due to neurological conditions, aging, or injuries. Accessing physical therapy is often difficult due to a shortage of therapists and other barriers to care. To address this issue, many patients rely on self-directed rehabilitation programs that allow them to exercise at home. While these programs can be beneficial, they often lack professional guidance, increasing the risk of incorrect exercise execution or loss of motivation. This project will enhance the effectiveness and engagement of home rehabilitation by leveraging artificial intelligence (AI) and motion-sensing technology. Beyond improving rehabilitation tools, the project will provide opportunities for students to work at the intersection of engineering and healthcare. Students will gain hands-on experience developing innovative technologies, exploring entrepreneurship, and engaging in public outreach to raise awareness of AI-powered rehabilitation solutions. This CAREER proposal focuses on advancing adaptive, self-directed hand rehabilitation programs through three technical innovations. First, it will develop a computer vision-based recovery monitoring system that integrates motion sensing and muscle activity data to model and visualize hand recovery dynamics. These recovery models will serve as real-time feedback to patients, offering a detailed understanding of their progress. Second, it aims to explore the predictive power of physiological signals and verbalized (think-aloud) data for adherence levels, establishing data-driven insights into patient behavior. Third, the project will design and evaluate an AI-supported rehabilitation program that leverages these insights to provide personalized care. A generative AI module will dynamically adapt multimedia interventions, offering rewards and encouragement based on adherence predictions, to enhance patient motivation and engagement. This work incorporates advanced methodologies, including motion analysis, machine learning, and generative AI, to create an integrated system that bridges recovery monitoring and motivational support. The findings will lead to fundamental advancements in recovery dynamics modeling and adaptive intervention strategies, paving the way for sustainable, tailored self-directed rehabilitation solutions. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.