Desenvolvimento de uma plataforma multimodal para treinamento de próteses de membro superior

Abstract

The lack of a upper limb is critical while performing daily tasks, and can directly affect the subject's quality of life. For this reason, interest has been growing in the development of increasingly robust prosthesis. However, rejection rates are still high, since prosthesis control still remains a challenge, especially in the early stages. In order to solve this problem, strategies that use virtual environments as training platforms have been developed, allowing the subject to focus on the prosthesis control. This work presents two forms of training in a virtual environment for two types of the most common prosthesis in the market: the body-powered prosthesis and the myoelectric prosthesis. Both systems also have an inertial sensor in order to track the movement performed by the limb and to increase the feedback offered by the virtual environment. To validate the implemented system, tests were performed and evolution was observed according to the number of times the user dropped the object as well as the time is taken to complete the task. It was noticed that there was an advance for all volunteers, especially when considering the number of objects lost, showing that the implemented system can be used as a tool to support training therapies for controlling upper limb prosthesis.