Design and Interaction Optimization of Smart Hand Rehabilitation Assistive Devices for Stroke Patients

Abstract

Long term disability following a stroke is complicated by hand dysfunction which is a major contributor and is a significant impediment to patients' independence, and quality of life. In this meta-analysis, the study analyzes the effects of design and interaction optimization in smart hand rehabilitation assistive devices for stroke patients. The analyses of a random effects model to systematically reviewed and analyzed 26 peer-reviewed studies published between 2015 and 2025 were, in total, conducted. The devices included robotic exoskeletons, wearable smart gloves, EMG based tools, and AI-driven adaptive systems.

Results indicate that smart devices have a major impact on hand function and AI-based systems showed the highest effect size (SMD = 0.92). Gamified and interactive elements (tasks, real-time feedback, and adaptive interface) improved the motor learning and patient engagement and improved therapy duration and faster recovery timeline. Not only this but the combination of intelligent systems in the rehabilitation process reduced therapist contact hours and reimbursements expenditure as a whole, resulting on average a saving of $1,524 per patient in six weeks.

Although differences in device types and patient demographics were observed, no differences were found in the evidence for the superiority of (optimized) smart rehabilitation technologies over conventional methods. These findings demonstrate that computational design coupled with real-time interaction and personalized feedback enables faster post-stroke recovery. Based on the work presented, the study concludes with a call for longitudinal studies with an interdisciplinary perspective to evolve intelligent rehabilitation solutions.