| Intelligent Home-Based Monitoring and Evaluation of Corrective Exercises Using Artificial Intelligence: A Narrative Review |
| Paper ID : 1132-SPORTCONGRESS (R1) |
| Authors |
|
Mohamad ali Biabangard, Saeid Azadbakht *, Ali asghar Maleki Department of Sport Injuries and Biomechanics, Faculty of Sport Sciences and Health, University of Tehran |
| Abstract |
| Abstract Introduction: Correct execution of corrective exercises is essential for improving posture and preventing musculoskeletal injuries. However, many patients and athletes perform these exercises incorrectly at home due to lack of real-time expert feedback. Advances in computer vision and artificial intelligence have enabled intelligent systems that automatically monitor and evaluate exercise performance using standard cameras or motion sensors. These systems can reduce the need for in-person supervision and improve adherence to rehabilitation programs. This study reviews recent research on AI-based monitoring of home-based corrective exercises. Methods: This study was conducted as a narrative review. Relevant literature was searched in PubMed, Scopus, and Google Scholar between 2020 and 2025 using keywords such as “corrective exercise monitoring,” “artificial intelligence,” “pose estimation,” and “home rehabilitation.” The selected studies were evaluated based on their scientific quality and conceptual relevance and were analyzed narratively. Results: The reviewed studies showed that computer vision-based pose estimation algorithms can detect joint angles and movement patterns in real time with reported accuracies ranging from 80% to 95% in controlled environments. Most systems relied on 2D RGB video input from standard webcams, while a few used depth sensors or wearable inertial units. These systems can provide real-time visual or auditory feedback to correct performance errors and, in some cases, track user progress over time. However, challenges remain, including the need for more diverse training datasets, reduced accuracy in uncontrolled home environments, and concerns about data privacy. Conclusion: AI-based intelligent systems have strong potential as complementary tools to monitor and enhance corrective exercise performance at home. Future research should aim to develop more accurate algorithms, create user-friendly interfaces for non-experts, and conduct large-scale studies to assess clinical effectiveness. Ultimately, these systems could make home rehabilitation more accessible, personalized, and scalable, reducing demands on clinical resources and supporting long-term musculoskeletal health improvements. |
| Keywords |
| Artificial Intelligence; Pose Estimation; Corrective Exercise; Movement Monitoring |
| Status: Abstract Accepted (Poster Presentation) |
