Sub135 SYSTEMATIC REVIEW ON THE ROLE OF ARTIFICIAL INTELLIGENCE IN DIAGNOSING AND MANAGING MUSCULOSKELETAL DISORDERS AND INJURIES

Résumé fourni par la source

Background: Musculoskeletal disorders represent a leading cause of global disability, creating a pressing need for innovations in diagnostic and management strategies. Artificial intelligence (AI) has emerged as a transformative tool with potential applications in interpreting medical images and predicting patient outcomes for these conditions. However, the evidence regarding its efficacy remains fragmented across various applications, necessitating a comprehensive synthesis. Objective: This systematic review aims to evaluate the current evidence on the performance of AI applications in diagnosing and managing a broad spectrum of musculoskeletal disorders and injuries, comparing its accuracy to standard clinical or radiological assessments. Methods: A systematic literature search was conducted in accordance with PRISMA guidelines across PubMed/MEDLINE, Scopus, Web of Science, and the Cochrane Library for studies published between January 2014 and June 2024. Inclusion criteria encompassed original studies evaluating AI models for musculoskeletal conditions against a reference standard. Two independent reviewers performed study selection, data extraction, and risk of bias assessment using appropriate tools like QUADAS-2. A qualitative synthesis was undertaken due to heterogeneity. Results: Eight studies involving 21,450 patients and image series were included. AI models, primarily deep learning networks, demonstrated high performance in detecting fractures (AUC up to 0.97, sensitivity up to 98.5%), grading osteoarthritis (accuracy up to 88%), and assessing soft-tissue pathologies like meniscal and rotator cuff tears (sensitivity 94%, correlation r=0.91). Performance was often statistically significant (p<0.001) and comparable to expert clinicians. Common limitations in the included studies were retrospective design and potential for verification bias. Conclusion: AI models show significant promise as accurate tools for assisting in the diagnosis and quantification of musculoskeletal conditions, performing on par with clinical experts in controlled research settings. These findings support their potential role as decision-support tools in clinical workflows. Future research should prioritize prospective, real-world validation and focus on evaluating the impact of AI integration on long-term patient outcomes and clinical efficiency.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.