Design And Development Of Bionic 3d Printed Upper-Limb Exoskeleton For Stroke Rehabilitation

Abstract DESIGN AND DEVELOPMENT OF BIONIC 3D PRINTED UPPER-LIMB EXOSKELETON FOR STROKE REHABILITATION Those who have had a stroke need long-term physical rehabilitation treatments. The conventional manual hands-on method is time-consuming and cannot offer low-cost, long-term rehabilitation services. Furthermore, the hands-on method is subjective and is dependent on the competence level of physiotherapists. To address these problems, more research and effort are being directed into the development of robotic-assisted rehabilitation technologies. For decades, upper-limb exoskeleton devices have been researched and developed for upper-limb rehabilitation. Some products have been tested on post-stroke patients. However, there are many limits and problems in this field. It currently needs a bionic mechanical design for the upper-limb exoskeleton, focusing mainly on human-machine joint misalignment. Furthermore, upper-limb exoskeleton volitional control is still far from acceptable.In this study, previous attempts to develop an upper-limb exoskeleton were discussed. Important challenges related to mechanical exoskeleton design and control strategy development were undertaken. A new robot control mechanism, known as the mirror or synchronous motion-control method, was developed, as well as tactics for promoting robotic systems in clinical usage. Six healthy individuals participated in an experiment to verify the performance of the motion-controlled upper-limb (UL) exoskeleton. In a 2D panel, the UL exoskeleton executed sketching movements based on the UL movements of a healthy participant. To evaluate the accuracy of the drawing performance, the drawings produced by the UL exoskeleton were compared to drawings done by the participant. To evaluate the inter-rater agreement between the drawings, and thus evaluate UL exoskeleton performance, a reliability statistical study (Cronbach test) was conducted and an image comparison using Python was performed to see the accuracy