Signal processing for remote monitoring of home-based rehabilitation support activities

Abstract

In the field of tele-rehabilitation, to better analyze a patient’s performance during motor activity, obtained through biomedical instrumentation and/or digital technologies, it is necessary to process and evaluate the signals extracted from the various sensors employed. In the literature there is often a lack of in-depth studies regarding such approaches and methodologies of signal processing. The purpose of the present paper is to define possible approaches for processing raw signals, paying attention to the type of noise that can afflict the signal. In the application of the ReMoVES IoT system, a procedure is here proposed and applied to analyze the signals coming from the Microsoft Kinect sensor, which is used to detect upper limb movements while performing the exercise, in order to study the behavior of a group of healthy subjects, and compare it with the performance of a patient who first performed a training phase in an inpatient setting and then for about a month, a treatment plan at home.