Objective Motion Tracking for Rehabilitation: Quantitative Assessment Using Pose Estimation

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Evaluation of the effectiveness of rehabilitation therapy is often based on patient outcomes and functional improvement over time. However, traditional approaches are subjective and heavily influenced by expert perspectives and perceptions, which can lead to inconsistencies in research. This study developed a motion measurement system that improves body positioning by using computer vision techniques to quantitatively assess elbow flexibility and range of motion during rehabilitation. The system uses a novel posture estimation model that analyses video input to identify key body points and match body position, providing smooth, real-time feedback. In addition, the system is designed to be easy to use and simple, so it can be implemented in both clinical settings and home-based rehabilitation programs. Data validation is carried out using a reference dataset that has previously been analyzed, with an absolute mean error of approximately 5.00 degrees and a correlation coefficient of approximately 0.95. The results show that this system can be used to maintain one's own behavior. This technology supports the rehabilitation process by providing clear data, helping medical professionals and health care providers better assess patient progress and make more appropriate decisions. In addition, this technology can improve tele-rehabilitation services, especially for patients who cannot access medical facilities. With that in mind, limitations such as model issues and video quality variations can affect the accuracy of body-position determination.