Use of Augmented Reality and Artificial Intelligence in Rehabilitation (Atena Editora)

Abstract

The regulation of telerehabilitation in Brazil, due to the Covid-19 pandemic, was made official by COFFITO Resolution No. However, in the absence of telerehabilitation facilitating devices, there is a need to develop tools that would enable the prescription, follow-up and monitoring of sessions. The objective of this work is to evaluate the system developed at the IFRJ that uses the Kinect device and its functionalities to monitor the execution of exercises prescribed by a professional physiotherapist or occupational therapist. The methodology used involves the construction of an exercise monitoring and prescription software by the multidisciplinary team (IFRJ Innovation) to work in the Teleconsultation and Telemonitoring system. A protocol was elaborated to standardize the correct execution of the movement and, later, tests were carried out to evaluate the system's ability to identify errors, comparing them to the established standard movement. The software was used Make Human and Blender, where it was possible to create an animated guide for the correct execution of the exercise. We opted for the “squatting” movement for the first phase of the program, with observation in an anterior view and identification of possible errors: dynamic knee valgus, anterior inclination (flexion) of the trunk and asymmetrical movements of the lower limbs. During the execution of the movement, the patient is monitored by the Kinect that supplies the data to an artificial intelligence network in order to obtain an instantaneous feedback of the correct or incorrect execution. Initial tests demonstrated that the software is capable of identifying the correct or wrong pattern of movements, but new exercises/protocols must be studied and tested in order to improve the system.