An Analysis of the Knee Injury Rehabilitation Via a Mobile-Computing Approach

Abstract

Major knee injuries and problems often occur during accidents, recreational activities or sports. Depending on the severity, a knee injury typically takes a long time to recover. Therefore, in order to promote knee recovery, knee exercises have proven to be very crucial and important to build strength and recover the range of motion of the injured knee. However, since the cost of rehabilitation is usually high, most patients would opt to complete it on their own at home. Also, as the rehabilitation protocol is complex by nature, it is very challenging for a patient to accomplish it without any professional guidance. As a result, many patients will not be able to fully recover from the injury. Thus, we have proposed an effective and low-cost approach to overcome this problem. In this project, we used the sensors, i.e., accelerometer and gyroscope, in a smartphone to collect the knee rotation data, and used machine learning techniques, particularly artificial neural network, to analyze the collected data. The goal is to provide an effective solution to help patients achieve effective rehabilitation.