MOTION CAPTURE AND ANTERIOR CRUCIATE INJURY RISK: IS THERE A GOLD STANDARD? A SYSTEMATIC REVIEW

Le résumé fourni par la source

The anterior cruciate ligament (ACL) plays a critical role in knee stabilisation during dynamic movement but is highly susceptible to injury, particularly in female athletes, who face a 2–10 times greater risk of non-contact ACL injuries than males. Injury prevention programs (IPPs) can effectively reduce injury incidence, demonstrating the importance of identifying high-risk individuals. While traditional motion capture technology shows promise for screening by analysing biomechanical deficits and movement patterns, its high cost and complexity limit clinical use, driving interest in accessible alternatives. This review examines whether a reliable gold standard, including a specific task and biometric pairing for 2D and 3D motion capture, has been established to evaluate such technologies. This systematic review adhered to PRISMA guidelines and was registered with PROSPERO (CRD42024602414). A comprehensive search of MEDLINE, EMBASE, COCHRANE, CINAHL, and AMED databases was conducted. Studies were included if they investigated motion capture technology in predicting non-contact ACL injury risk in adult female athletes (>18 years), published after January 1, 2000, and available in English. Two reviewers independently screened all identified studies, extracted data, and assessed the results. Data extraction encompassed: biomechanical variables (e.g. joint kinematics and kinetics); functional movement patterns, motion capture technology, study design and participant characteristics; and predictive metrics (e.g. sensitivity, specificity and predictive success). 9 studies, comprising 1402 participants (1392 female), met the criteria for inclusion. The review identified significant variability in the methodologies used, predictive validity, and reliability of motion capture. Predictive accuracy, measured via area under the curve (AUC), ranged from 0.42–0.803, while intraclass correlation coefficients (ICCs) for reliability varied from 0.22–0.99, with higher consistency observed within sessions (0.87–0.99) compared to between sessions across all task and biometric measurement pairings. Regarding task performance and biometric analysis, the drop-vertical jump combined with lateral trunk motion emerged as the most accurate measurement combination, though further research is needed to confirm these results. Heterogeneity in study designs, tasks, and motion capture technologies limited comparability across studies. This review found high heterogeneity in the current literature regarding the accuracy of different motion capture tasks and biometric measurement pairings for 2D and 3D studies. Therefore, it cannot be confidently concluded that the field has a reliable method for evaluating the accuracy of emerging motion capture techniques. This review demonstrates the need for improved standardisation and validation across study designs, tasks, and populations to enable the progression of motion capture as a predictive tool.

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