Sistema de rehabilitación física para pacientes con lesiones de tobillo empleando visión artificial para la Unidad de Rehabilitación y Fisioterapia Gabo’s

Abstract

This project is oriented to the implementation of a physical rehabilitation system for patients with ankle injuries using artificial vision in the Gabo's Rehabilitation and Physiotherapy Unit. For the virtual environment a Kinect V2.0 camera is used which captures the patient's movements and replicates them in the avatar that was developed in Unity, this avatar is presented on a monitor for the patient to observe their movements. The implemented system consists of a Jetson Nano minicomputer, which runs the Python program to analyze and process the patient's joint data. The angles formed by the three reference points are calculated and sent using the MQTT protocol to Node-Red, which is responsible for collecting the data and subsequently storing them in Influxdb. The database with the information of the ankle angles is generated, the respective configuration of Influxdb is made in Grafana, to observe the information stored in a graphical interface where the ankle angles generated by the ankles can be appreciated. The system implemented in the physiotherapy unit in comparison with the measurements obtained manually using a goniometer presents a reliability of 95.28% in total average, in obtaining the angles of the joint.