Black Box Prediction Methods in Sports Medicine Deserve a Red Card for Reckless Practice: A Change of Tactics is Needed to Advance Athlete Care

Head Exposure to Acceleration Database in Sport (HEADSport): a kinematic signal processing method to enable instrumented mouthguard (iMG) field-based inter-study comparisons

Objective

Instrumented mouthguard (iMG) systems use different signal processing approaches limiting field-based inter-study comparisons, especially when artefacts are present in the signal. The objective of this study was to assess the frequency content and characteristics of head kinematic signals from head impact reconstruction laboratory and field-based environments to develop an artefact attenuation filtering method (HEADSport filter method).

Methods

Laboratory impacts (n=72) on a test-dummy headform ranging from 25 to 150 g were conducted and 126 rugby union players were equipped with iMGs for 209 player-matches. Power spectral density (PSD) characteristics of the laboratory impacts and on-field head acceleration events (HAEs) (n=5694) such as the 95th percentile cumulative sum PSD frequency were used to develop the HEADSport method. The HEADSport filter method was compared with two other common filtering approaches (Butterworth-200Hz and CFC180 filter) through signal-to-noise ratio (SNR) and mixed linear effects models for laboratory and on-field events, respectively.

Results

The HEADSport filter method produced marginally higher SNR than the Butterworth-200Hz and CFC180 filter and on-field peak linear acceleration (PLA) and peak angular acceleration (PAA) values within the magnitude range tested in the laboratory. Median PLA and PAA (and outlier values) were higher for the CFC180 filter than the Butterworth-200Hz and HEADSport filter method (p<0.01).

Conclusion

The HEADSport filter method could enable iMG field-based inter-study comparisons and is openly available at https://github.com/GTBiomech/HEADSport-Filter-Method.

Players', Head Coaches', And Medical Personnels' Knowledge, Understandings and Perceptions of Injuries and Injury Prevention in Elite-Level Women's Football in Ireland

BACKGROUND

To manage injuries effectively, players, head coaches, and medical personnel need to have excellent knowledge, attitudes, and behaviours in relation to the identification of risk factors for injuries, the implementation of injury prevention initiatives, as well as the implementation of effective injury management strategies. Understanding the injury context, whereby specific personal, environmental, and societal factors can influence the implementation of injury prevention initiatives and injury management strategies is critical to player welfare. To date, no qualitative research investigating the context of injuries, has been undertaken in elite-level women's football. The aim of our study was to explore the knowledge, attitudes, and behaviours of players, head coaches, and medical personnel in the Irish Women's National League (WNL) to injury prevention and injury management.

METHODS

We used qualitative research methods to explore the knowledge, attitudes, and behaviours of players, head coaches, and medical personnel in the Irish WNL to injury prevention and injury management. Semi-structured interviews were undertaken with 17 players, 8 medical personnel, and 7 head coaches in the Irish WNL. The data were analysed using thematic analysis. Our study is located within an interpretivist, constructivist research paradigm.

RESULTS

The participants had incomplete knowledge of common injuries in elite-level football, and many held beliefs about risk factors for injuries, such as menstrual cycle stage, which lacked evidence to support them. Jumping and landing exercises were commonly used to reduce the risk of injuries but evidence-based injury prevention exercises and programmes such as the Nordic hamstring curl, Copenhagen adduction exercise, and the FIFA 11+ were rarely mentioned. Overall, there was dissatisfaction amongst players with their medical care and strength and conditioning (S & C) support, with resultant inadequate communication between players, head coaches, and medical personnel.

CONCLUSION

Poor quality and availability of medical care and S & C support were considered to be a major obstacle in the effective implementation of injury risk reduction strategies and successful return-to-sport practices. More original research is required in elite-level women's football to explore injury risk factors, injury prevention initiatives, and contextual return-to-sport strategies, so that players, head coaches, and medical personnel can use evidence that is both up-to-date and specific to their environment.

Relationship between a daily injury risk estimation feedback (I-REF) based on machine learning techniques and actual injury risk in athletics (track and field): protocol for a prospective cohort study over an athletics season

INTRODUCTION

Two-thirds of athletes (65%) have at least one injury complaint leading to participation restriction (ICPR) in athletics (track and field) during one season. The emerging practice of medicine and public health supported by electronic processes and communication in sports medicine represents an opportunity for developing new injury risk reduction strategies. Modelling and predicting the risk of injury in real-time through artificial intelligence using machine learning techniques might represent an innovative injury risk reduction strategy. Thus, the primary aim of this study will be to analyse the relationship between the level of Injury Risk Estimation Feedback (I-REF) use (average score of athletes' self-declared level of I-REF consideration for their athletics activity) and the ICPR burden during an athletics season.

METHOD AND ANALYSIS

We will conduct a prospective cohort study, called Injury Prediction with Artificial Intelligence (IPredict-AI), over one 38-week athletics season (from September 2022 to July 2023) involving competitive athletics athletes licensed with the French Federation of Athletics. All athletes will be asked to complete daily questionnaires on their athletics activity, their psychological state, their sleep, the level of I-REF use and any ICPR. I-REF will present a daily estimation of the ICPR risk ranging from 0% (no risk for injury) to 100% (maximal risk for injury) for the following day. All athletes will be free to see I-REF and to adapt their athletics activity according to I-REF. The primary outcome will be the ICPR burden over the follow-up (over an athletics season), defined as the number of days lost from training and/or competition due to ICPR per 1000 hours of athletics activity. The relationship between ICPR burden and the level of I-REF use will be explored by using linear regression models.

ETHICS AND DISSEMINATION

This prospective cohort study was reviewed and approved by the Saint-Etienne University Hospital Ethical Committee (Institutional Review Board: IORG0007394, IRBN1062022/CHUSTE). Results of the study will be disseminated in peer-reviewed journals and in international scientific congresses, as well as to the included participants.

Predictive modeling of the ratings of perceived exertion during training and competition in professional soccer players

OBJECTIVES

Evaluate the ability of predicting the ratings of perceived exertion from the external load variables in professional soccer players through a chronological perspective (i.e., past features values are considered additional features) through machine learning models by considering the playing position.

DESIGN

Prospective cohort study.

METHODS

Thirty-eight elite soccer players aged 19-27 years were observed during 151 training sessions, 44 matches across a full season. External load variables (58 derived from Global Positioning System and 30 from accelerometers) and the internal load derived from ratings of perceived exertion were collected for each player and each session and match. Machine learning models (linear regression, K-NN, decision trees, random forest, elastic net regression, XGBoost) were compared and interpreted in order to deepen the relationship between external load variables and ratings of perceived exertion according to the player position in a predictive perspective.

RESULTS

Application of the machine learning models on the dataset provided enough predictive power to reduce the Root Mean Squared Error of 60 % from dummy predictions. The most accurate models (Root Mean Squared Error ≈ 1.1 for random forest and = 1 for XGBoost) highlighted a memory effect in subsequent ratings of perceived exertion values. Past ratings of perceived exertion values over one month were the strongest predicting factors of ratings of perceived exertion as compared to various external load indicators.

CONCLUSIONS

The tree-based machine learning models showed statistically significant predictive ability, indicating valuable information for understanding the training load responses based on ratings of perceived exertion changes.

Predictive utility of commercial grade technologies for assessing musculoskeletal injury risk in US Marine Corps Officer candidates

Recently, commercial grade technologies have provided black box algorithms potentially relating to musculoskeletal injury (MSKI) risk and functional movement deficits, in which may add value to a high-performance model. Thus, the purpose of this manuscript was to evaluate composite and component scores from commercial grade technologies associations to MSKI risk in Marine Officer Candidates. 689 candidates (Male candidates = 566, Female candidates = 123) performed counter movement jumps on SPARTA™ force plates and functional movements (squats, jumps, lunges) in DARI™ markerless motion capture at the start of Officer Candidates School (OCS). De-identified MSKI data was acquired from internal OCS reports for those who presented to the Physical Therapy department for MSKI treatment during the 10 weeks of training. Logistic regression analyses were conducted to validate the utility of the composite scores and supervised machine learning algorithms were deployed to create a population specific model on the normalized component variables in SPARTA™ and DARI™. Common MSKI risk factors (cMSKI) such as older age, slower run times, and females were associated with greater MSKI risk. Composite scores were significantly associated with MSKI, although the area under the curve (AUC) demonstrated poor discrimination (AUC = .55-.57). When supervised machine learning algorithms were trained on the normalized component variables and cMSKI variables, the overall training models performed well, but when the training models were tested on the testing data the models classified MSKI "by chance" (testing AUC avg = .55-.57) across all models. Composite scores and component population specific models were poor predictors of MSKI in candidates. While cMSKI, SPARTA™, and DARI™ models performed similarly, this study does not dismiss the use of commercial technologies but questions the utility of a singular screening task to predict MSKI over 10 weeks. Further investigations should evaluate occupation specific screening, serial measurements, and/or load exposure for creating MSKI risk models.

Estimating Ground Reaction Forces from Two-Dimensional Pose Data: A Biomechanics-Based Comparison of AlphaPose, BlazePose, and OpenPose

The adoption of computer vision pose estimation approaches, used to identify keypoint locations which are intended to reflect the necessary anatomical landmarks relied upon by biomechanists for musculoskeletal modelling, has gained increasing traction in recent years. This uptake has been further accelerated by keypoint use as inputs into machine learning models used to estimate biomechanical parameters such as ground reaction forces (GRFs) in the absence of instrumentation required for direct measurement. This study first aimed to investigate the keypoint detection rate of three open-source pose estimation models (AlphaPose, BlazePose, and OpenPose) across varying movements, camera views, and trial lengths. Second, this study aimed to assess the suitability and interchangeability of keypoints detected by each pose estimation model when used as inputs into machine learning models for the estimation of GRFs. The keypoint detection rate of BlazePose was distinctly lower than that of AlphaPose and OpenPose. All pose estimation models achieved a high keypoint detection rate at the centre of an image frame and a lower detection rate in the true sagittal plane camera field of view, compared with slightly anteriorly or posteriorly located quasi-sagittal plane camera views. The three-dimensional ground reaction force, instantaneous loading rate, and peak force for running could be estimated using the keypoints of all three pose estimation models. However, only AlphaPose and OpenPose keypoints could be used interchangeably with a machine learning model trained to estimate GRFs based on AlphaPose keypoints resulting in a high estimation accuracy when OpenPose keypoints were used as inputs and vice versa. The findings of this study highlight the need for further evaluation of computer vision-based pose estimation models for application in biomechanical human modelling, and the limitations of machine learning-based GRF estimation models that rely on 2D keypoints. This is of particular relevance given that machine learning models informing athlete monitoring guidelines are being developed for application related to athlete well-being.

Forecasting football injuries by combining screening, monitoring and machine learning

Identifying players or circumstances associated with an increased risk of injury is fundamental for successful risk management in football. So far, time-constant and volatile risk factors are generally considered separately in either a screening (constant) or a monitoring (volatile) approach each resulting in a restricted set of explanatory variables. Consequently, improvements in predictive accuracy may be expected when screening and monitoring data are combined, especially when analysed with current machine learning (ML) techniques. This trial was designed as a prospective observational cohort study aiming to forecast non-contact time-loss injuries in male professional football (soccer). Injuries were registered according to the Fuller consensus. Gradient boosting with ROSE upsampling within a leave-one-out cross-validation was used for data analysis. The hierarchical data structure was considered throughout. Different splits of the original dataset were used to probe the robustness of results. Data of 88 players from 4 teams and 51 injuries could be analysed. The cross-validated performance of the gradient boosted model (ROC area under the curve 0.61) was promising and higher compared to models without integration of screening data. Importantly, holdout test set performance was similar (ROC area under the curve 0.62) indicating prospect of generalizability to new cases. However, the variation of predictive accuracy and feature importance with different splits of the original dataset reflects the relatively low number of events. It is concluded that ML-based injury forecasting based on the integration of screening and monitoring data is promising. However, external prospective verification and continued model development are required.