Kinect based system and Artificial Neural Networks classifiers for physiotherapy assessment

This paper describes the development of a health monitoring platform for physiotherapy patients, particularly focused on the ones suffering from stroke disease. Using the Kinect sensor, the presented system intends to provide technical data about the patients that could help therapists during physiotherapy assessments. For helping quantifying the patients´ performance, a mechanism for human poses and gestures recognition was built using a sequence of Artificial Neural Networks (ANNs). Furthermore, through remote physiotherapy our system helps ensuring the continuity of the treatments outside the clinics and a considerable cost reduction.