Volume - 4 | Issue - 4 | december 2022
DOI
10.36548/jeea.2022.4.003
Published
29 November, 2022
The work presented in this paper addresses the enhancement of upper body rehabilitation and training methods for stroke victims and upper body amputees. One of the primary aims is to develop a tool that utilizes augmented reality to facilitate the rehabilitation of impaired human hand and forearm movements by employing mirror neurons and virtual reality. The second objective of the proposed tool is to allow for evaluation and specification of prostheses prior to acquisition and fitting of such devices to upper limb amputees. The proposed system involves the development of real–time surface Electromyography (sEMG) signal classification methods, Artificial Neural Network training, and implementation and the development of identification algorithms for inferring motion intend. The results of the proposed approach indicate preferences of specific classifiers used in the processing of sEMG data. The proposed methods are implemented in a virtual reality environment allowing for potential selection and training of prosthetic device usage as well as for physical therapy rehabilitation sessions of stroke victims.
KeywordsArtificial neural network motion identification rehabilitation prosthesis hand real – time model surface electromyography classification virtual reality
