Smart solution for pain detection in remote rehabilitation

Abstract

In this article, we present a low cost open source robotic solution for remote rehabilitation. In a pandemic context, where the trend is to manage our services remotely, the proposed system guarantees that the physiotherapist remotely controls and supervises the rehabilitation process. In the developed solution, it is not necessary to directly touch either the subject or the system, since the proposed architecture ensures control of the device remotely and using only a simple gesture. According to experts, pain is considered one of the major constraints that prevent patients from easily accepting rehabilitation sessions. To our knowledge, there is no method or tool presented in the literature to measure pain. The proposed solution aims to estimate the pain then to manage the control of the robot according to this estimate. In fact, the idea is to estimate pain through a decision support system based on cascading fuzzy logic. The first compartment of this fuzzy system was used to estimate muscle contact based on the EMG signal. The second compartment was set up to estimate pain according to three parameters: muscle contraction, the patient's resistance force and the last angle reached. A user-friendly human machine interface (HMI) has been developed to ensure communication between the robot and the physiotherapist. The control is done by gestures taking advantage of the Kinect technology. Sensory information about strength, range of motion (RoM) and estimated pain level is communicated to the physiotherapist's board. Patient information, health changes and tracked activities are all recorded, providing an electronic health record. Experimental studies carried out on a set of patients in a clinical environment have shown the efficiency and reliability of the developed robotic solution.