Prediction of injuries in elite soccer players with the analysis of asymmetries in the CMJ through the use of Machine Learning tools

Introduction: The analysis of asymmetries in the countermovement jump (CMJ) is a potentially useful tool for the prevention of injuries in elite soccer players. Objective: This study analyzes the relationship between biomechanical asymmetries in the CMJ on landing (% Peak Landing Force Assymetry) and the risk of musculoskeletal injuries. Methodology: Through the use of force platforms (ForceDecks, Valdperformance) and 4 machine learning models, data from 29 Asian Football Confederation (AFC) Champions League elite level professional soccer players were analyzed during a regular season (with a total of 1265 jumps analyzed, during the days Match Day Training MD+1, MD+2 and MD-1). Discussion: The results indicate that the average asymmetries (0.82% +- 13.66) in the peak landing force are not predictive of the risk of injury, while the % Coefficient of variation (COV) (323–1827) of that asymmetry was related and could have a predictive nature regarding the risk of injuries. Furthermore, there is a moderate positive correlation (r = 0.447) between % COV Assymetry and the appearance of injuries, which is statistically significant (p = 0.017). Conclusions: The odds ratio of 2.4 indicates that players with higher than average COV asymmetry are 2.4 times more likely to suffer an injury. These findings highlight the importance of the CMJ as part of injury monitoring and injury reduction programs.