Evaluation of Upper Limb Joint's Range of Motion Data by Kinect Sensor for Rehabilitation Exercise Game

Abstract

Rehabilitation is a tedious process both for patients and physiotherapists. Many games have been developed, especially for upper limb rehabilitation, to help training more enjoyable for patients which results in an increase in participation and shortening recovery time. Using game, patient's performance in each training course is assessed by the score patient receives after the session. Although these scores might correlate to the range of motion (ROM) of the arm, they cannot be used to directly for evaluation. In this study, a game was created for upper limb rehabilitation which also provides summary of upper limp movement over training session in terms of ROM by using Kinect camera. Formulae to obtain seven upper limb's ROM angles i.e., especially for shoulder flexion, shoulder extension, shoulder abduction, shoulder adduction, shoulder internal rotation, shoulder external rotation and elbow flexion, have been derived. Experiments were conducted to evaluate accuracy of the ROM obtained from the calculation with respect to conventional measurement using goniometer. The results show that, with proper setting (position and orientation) of the Kinect sensor, ROM values obtained from the Kinect sensor match well to the goniometer to within +/-13.5%. Various statistics related to ROM data can be analyzed after each training session. This game would be a useful tool for physiotherapist to monitor and evaluate the progress of patient's recovery.