Abstract:
This project aims to develop an efficient, cost-effective, and biocompatible sEMG driven Standalone Rehabilitation System for the assistive bilateral training of stroke patients with compromised upper limb functionality. The proposed system comprises of an Assistive Soft-Robotics Smart Wearable Glove and a Rehabilitation Protocol that gamifies the traditional rehabilitation exercises for providing an interactive experience to stroke patients. Multiple Ensemble learning algorithms (i.e., XGBoost, Random Forest, Gradient Boost and Decision Tress.) were compared for the real time classification of sEMG data extracted from the functional hand of stroke patients using Mayo armband. The proposed Soft Robotic Glove will utilize surface Electromyography (sEMG) signals and flexible mechanics to increase robustness and safety in grasping and manipulation strategies. The master-slave control was developed for the actuation of assistive glove to ease bilateral training and was integrated in such a way that a gesture of the functional hand translates to the movement in parietal hand via the soft robotic glove. The project aimed to develop an effective, biocompatible, interactive, and patient-directed standalone rehabilitation system for stroke patients,
A two-months long rehabilitation protocol was designed that comprised of multiple interactive games with integration of different real time and offline performance parameters (extracted from the Fitts’ Law) for the evaluation of stroke patient’s conditions overtime. A two-month long study was conducted on multiple stroke patients with upper limb functional impairment in MAYO Hospital Lahore under the supervision of professional physiotherapists. The results from the study showed a significant improvement in the paretic hand movement of stroke patients. The proposed standalone rehabilitation has the potential to revolutionize neuromuscular rehabilitation, providing patients with more engaging, effective, and personalized therapies.
