New Machine Learning Approach for Detection of Injury Risk Factors in Young Team Sport Athletes

Abstract

The purpose of this article is to present how predictive machine learning methods can be utilized for detecting sport injury risk factors in a data-driven manner. The approach can be used for finding new hypotheses for risk factors and confirming the predictive power of previously recognized ones. We used three-dimensional motion analysis and physical data from 314 young basketball and floorball players (48.4% males, 15.72±1.79 yr, 173.34±9.14 cm, 64.65±10.4 kg). Both linear (L1-regularized logistic regression) and non-linear methods (random forest) were used to predict moderate and severe knee and ankle injuries (N=57) during three-year follow-up. Results were confirmed with permutation tests and predictive risk factors detected with Wilcoxon signed-rank-test (p<0.01). Random forest suggested twelve consistent injury predictors and logistic regression twenty. Ten of these were suggested in both models; sex, body mass index, hamstring flexibility, knee joint laxity, medial knee displacement, height, ankle plantar flexion at initial contact, leg press one-repetition max, and knee valgus at initial contact. Cross-validated areas under receiver operating characteristic curve were 0.65 (logistic regression) and 0.63 (random forest). The results highlight the difficulty of predicting future injuries, but also show that even with models having relatively low predictive power, certain predictive injury risk factors can be consistently detected.

Key words

sports medicine - predictive methods - machine learning - knee injuries - ankle injuries - basketball and floorball

• References

• 1 Jacobsson J, Timpka T, Kowalski J. et al. Prevalence of musculoskeletal injuries in Swedish elite track and field athletes. Am J Sports Med 2012; 40: 163-169
• 2 Emery CA, Rose MS, McAllister JR. et al. A prevention strategy to reduce the incidence of injury in high school basketball: a cluster randomized controlled trial. Clin J Sport Med 2007; 17: 17-24
• 3 Myklebust G, Holm I, Mæhlum S. et al. Clinical, functional, and radiologic outcome in team handball players 6–11 years after anterior cruciate ligament injury: a follow-up study. Am J Sports Med 2003; 31: 981-989
• 4 Bahr R, Holme I. Risk factors for sports injuries − a methodological approach. Br J Sports Med 2003; 37: 384-392
• 5 Murphy DF, Connolly DAJ, Beynnon BD. Risk factors for lower extremity injury: A review of the literature. Br J Sports Med 2003; 37: 13-29
• 6 Rossi A, Pappalardo L, Cintia P. et al. Effective injury forecasting in soccer with GPS training data and machine learning. PLoS One 2018; 13: e0201264
• 7 Bahr R. Why screening tests to predict injury do not work — and probably never will…: A critical review. Br J Sport Med 2016; 50: 776-780
• 8 Shmueli G. To explain or to predict?. Stat Sci 2010; 25: 289-310
• 9 López-Valenciano A, Ayala F, Puerta JM. et al. A preventive model for muscle injuries: A novel approach based on learning algorithms. Med Sci Sports Exerc 2018; 50: 915-927
• 10 Meeuwisse WH, Tyreman H, Hagel B. et al. A dynamic model of etiology in sport injury: The recursive nature of risk and causation. Clin J Sport Med 2007; 17: 215-219
• 11 Bittencourt NFN, Meeuwisse WH, Mendonça LD. et al. Complex systems approach for sports injuries: Moving from risk factor identification to injury pattern recognition — narrative review and new concept. Br J Sports Med 2016; 50: 1309-1314
• 12 Robertson S. Improving load/injury predictive modelling in sport: The role of data analytics. J Sci Med Sport 2014; 18: 25-26
• 13 Combrisson E, Jerbi K. Exceeding chance level by chance: The caveat of theoretical chance levels in brain signal classification and statistical assessment of decoding accuracy. J Neurosci Methods 2015; 250: 126-136
• 14 Biran O, Cotton C. Explanation and justification in machine learning: A survey. In: IJCAI-17 workshop on explainable AI (XAI). International Joint Conference on Artificial Intelligence; 2017
• 15 Bellazzi R, Zupan B. Predictive data mining in clinical medicine: Current issues and guidelines. Int J Med Inform 2008; 77: 81-97
• 16 Ruddy JD, Shield AJ, Maniar N. et al. Predictive modeling of hamstring strain injuries in elite Australian footballers. Med Sci Sports Exerc 2018; 50: 906-914
• 17 Dower C, Rafeli A, Weber J, Mohamad R. An enhanced metric of injury risk utilizing Artificial Intelligence. In: Proceedings of the 13^th annual MIT SLOAN Sports Analytics Conference. 2019
• 18 Pasanen K, Rossi MT, Parkkari J. et al. Predictors of lower extremity injuries in team sports (PROFITS-study): A study protocol. BMJ Open Sport Exerc Med 2015; 1: e000076
• 19 Harriss DJ, MacSween A, Atkinson G. Ethical standards in sport and exercise science research: 2020 update. Int J Sports Med 2019; 40: 813-817
• 20 Leppänen M, Pasanen K, Kujala UM. et al. Stiff landings are associated with increased ACL injury risk in young female basketball and floorball players. Am J Sports Med 2017; 45: 386-393
• 21 Fuller CW, Molloy MG, Bagate C. et al. Consensus statement on injury definitions and data collection procedures for studies of injuries in rugby union. Br J Sports Med 2007; 41: 328-331
• 22 Boulesteix A-L, Janitza S, Kruppa J. et al. Overview of random forest methodology and practical guidance with emphasis on computational biology and bioinformatics. Wiley Interdiscip Rev. WIREs Data Mining Knowl Discov 2012; 2: 493-507
• 23 Carey DL, Ong K, Whiteley R. et al. Predictive modelling of training loads and injury in Australian football. Int J Comput. Sci Sport 2018; 17: 49-66
• 24 Breiman L. Random forests. Mach Learn 2001; 45: 5-32
• 25 Fawcett T. An introduction to ROC analysis. Pattern Recognit Lett 2006; 27: 861-874
• 26 Krstajic D, Buturovic LJ, Leahy DE. et al. Cross-validation pitfalls when selecting and assessing regression and classification models. J Cheminform 2014; 6: 10-25
• 27 Vanderlei FM, Bastos FN, Tsutsumi GYC. et al. Characteristics and contributing factors related to sports injuries in young volleyball players. BMC Res Notes 2013; 6: 415
• 28 Jones BH, Bovee MW, Harris JM. et al. Intrinsic risk factors for exercise-related injuries among male and female army trainees. Am J Sports Med 1993; 21: 705-710
• 29 Boden BP, Dean GS, Feagin JA. et al. Mechanisms of anterior cruciate ligament injury. Orthopedics 2000; 23: 573-578
• 30 Yamamoto T. Relationship between hamstring strains and leg muscle strength. A follow-up study of collegiate track and field athletes. J Sports Med Phys Fitness 1993; 33: 194-199
• 31 Beynnon BD, Renström PA, Alosa DM. et al. Ankle ligament injury risk factors: A prospective study of college athletes. J Orthop Res 2001; 19: 213-220
• 32 Woodford-Rogers B, Cyphert L, Denegar CR. Risk factors for anterior cruciate ligament injury in high school and college athletes. J Athl Train 1994; 29: 343-346
• 33 Boden BP, Torg JS, Knowles SB. et al. Video analysis of anterior cruciate ligament injury: Abnormalities in hip and ankle kinematics. Am J Sports Med 2009; 37: 252-259
• 34 Hewett TE, Myer GD, Ford KR. et al. Biomechanical measures of neuromuscular control and valgus loading of the knee predict anterior cruciate ligament injury risk in female athletes: A prospective study. Am J Sports Med 2005; 33: 492-501