Preliminary Results of a Motion-Based Interactive Game for Supporting Stroke Survivor Telerehabilitation

Abstract

Telerehabilitation is the delivery of rehabilitation services over the internet or telecommunication services. In the light of current lockdowns and quarantines due to the COVID-19 telerehabilitation can be considered as the second-best option after rehab centers for stroke patients. The purpose of this study is to investigate the in-home telerehabilitation system prototype which introduces a state-of-the-art telerehabilitation solution with two extra features that are not currently available with other solutions; a communication module that connects the patient and the therapist remotely using a client-side application and an admin-side dashboard and the second feature is the variable motion-range setter, this option allows setting a certain percentage of a whole range of motion of an exercise based on the condition of the patient. The prototype application has 8 exercises targeting different parts of the body such as the ankle, hip, knee, and shoulder. With a controlled group of healthy male participants, a pilot study was designed to test the accuracy, error rate, recall, and other matrices of the developed system. The testing consisted of 50 repetitions of each exercise where 40 reps were performed as they should be performed, and the last 10 reps were performed to detect false positives. A confusion matrix was used to evaluate the performance of each exercise and the average of all the matrix is taken to get an overall performance result for the prototype. A sample of the results is included as the summary of results. The overall outcome is mostly positive and promising for most exercises, except for one exercise which was consistently on the lower spectrum of accuracy. After repositioning the Kinect to get the full body in the frame, the system performed consistently well on all exercises.