Design and Evaluation of an Exergame System of Knee with the Azure Kinect

Timely and effective knee function evaluation and knee exercises promote the prevention and self-management of knee diseases. In this paper, a Kinect-based exergame system is proposed to assess and train the knee function. Azure Kinect was used to capture and generate 3D models of the user and immerse them in an interactive virtual environment. The software included three functional modules: knee function evaluation, Knee exercises game, and Comprehensive evaluation. The stand, step, leg lift, and squat were selected for knee function evaluation and exercises. Twenty volunteers participated in the experiment. Intra-class correlation coefficients (ICCs) were calculated to assess the reliability of kinematic measurements of knee angles during the movements. The ICC of these movements were stand (ICC = 0.987), step (ICC = 0.997), left leg lift (ICC = 0.981), right leg lift (ICC = 0.990), stand (ICC = 0.998). The results show that the test-retest reliability is high. It means that the motion capture data is effective and the data obtained by Kinect is stable.