Use of Multilayer Perceptron vs. Distance Measurement Methods for Classification of Exercises in Telerehabilitation

Abstract

Abstract. Any effort to improve the efficiency of the physical rehabilitation processes is fundamental to ensure the sustainability of healthcare services. This efficiency depends greatly on the patient's adherence to the rehabilitation treatments. Information and communication technologies can help in these issues offering solutions that aim to monitor the patients' rehabilitation exercises performance allowing the existence of domiciliary rehabilitation scenarios. We have developed a solution of this kind, which aims to be as simple and low-cost as possible in the way of how recognizes and evaluates patient's movements. In this work we show a comparison between the use of a multilayer perceptron and a distance between patterns measuring algorithm for patients' motion recognition.

Physical Exercise Form Correction Using Neural Networks

“…Other works require additional hardware. [8] focuses on rehabilitation exercises and allows patients to continue their recovery from home. The technique requires a Nintendo Wii Remote © (motion sensor that communicates via Bluetooth) to be held in the hand, or attached to the moving limb.…”