Range of Motion Detection System in Humans Based on MediaPipe Special Flexion-Extension and Abduction-Adduction Movements

This research offers an alternative for health workers to monitor the rehabilitation of a patient's Range of Motion (ROM) without using a goniometer instrument and attaching various sensors to the patient's body, based on a detection system using the MediaPipe library. The developed detection system measures angles concerning the patient's body ROM patterns. The algorithm in the MediaPipe library helps to produce line graphs of landmark pose which are then processed using the trigonometric gradient equation so that it can have to measure angles from ROM. The angle of ROM measured includes Flexion-Extension and Abduction-Adduction movements of the arms, legs, and body. The results of testing the detection system between the goniometer measurements and the detection results differ from the Right Arm Elbow Flexion-Extension movement to the Trunk Flexion-Extension movement having an angle difference. The most significant angle difference is in the Right Arm Elbow Abduction - Adduction movement of 9.1388°, and the minor difference is in the Right Leg Hip Abduction - Adduction movement of 2.15°. The most significant overall system error is in the Left Arm Elbow Flexion - Extension movement, with a general system error of 27%. The minor system error in Left Arm Elbow Abduction - Adduction movement is 11 %.