Development of a Haptic Exo-Suit with Tactile Feedback in Hand Rehabilitation

Abstract

Stroke survivors with hemiplegia or paralysis often struggle with daily tasks due to impaired hand function and timely physiotherapy is essential for motor function recovery. This study presents a haptic forearm exo-suit designed to assist stroke rehabilitation by mimicking a physiotherapist's tactile feedback. The exo-suit integrates force-myography (FMG) and flexible tactile sensors to monitor muscle activity, finger movements, arm rigidity and tactile force in real-time. An AI-driven algorithm processes this sensor data to control the exo-suit's actuation system, which mimics natural finger movements and provides targeted therapeutic exercises. Initial testing demonstrated the exo-suit's effectiveness in simulating a physiothera-pist's precise and gradual actuation, improving the rehabilitation process. Evaluations on both healthy participants and those under simulated paralysis confirmed the hardware's capability and the reliability of the signal processing algorithms. Among the machine learning algorithms applied, k-nearest neighbors (KNN) achieved the highest accuracy at 86 %, with 89% precision and 84% recall. The proposed exo-suit offers an affordable and adaptive solution for individuals with partial arm paralysis, with real-time therapy adjustments that may significantly enhance recovery and improve the quality of life for stroke patients. © 2024 IEEE.