1
|
Elstak I, Salmon P, McLean S. Artificial intelligence applications in the football codes: A systematic review. J Sports Sci 2024; 42:1184-1199. [PMID: 39140400 DOI: 10.1080/02640414.2024.2383065] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/29/2023] [Accepted: 07/15/2024] [Indexed: 08/15/2024]
Abstract
Artificial Intelligence (AI) is increasingly being adopted across many domains such as transport, healthcare, defence and sport, with football codes no exception. Though there is a range of potential benefits of AI, concern has also been expressed regarding potential risks. An important first step in ensuring that AI applications in football are usable, beneficial, safe and ethical is to understand the current range of applications, the AI models adopted and their proposed functions. This systematic review aimed to identify different applications of AI across football codes to synthesise current knowledge and determine whether potential risks are being considered. The systematic review included 190 peer-reviewed articles. Nine areas of application were found ranging from athlete evaluation and event detection to match outcome prediction and injury detection and prediction. In total, 27 different AI models were identified, with artificial neural networks the most frequently applied. Five AI assessment metrics were identified including specificity, recall, precision, accuracy and F1-score. Four potential risks were identified, concerning data security, usability, data biases and inappropriate athlete load management. It is concluded that, though a wide range of AI applications currently exist, further work is required to develop AI for football and identify and manage potential risks.
Collapse
Affiliation(s)
- Isaiah Elstak
- Centre for Human Factors and Sociotechnical Systems, University of the Sunshine Coast, Sippy Downs, Queensland, Australia
| | - Paul Salmon
- Centre for Human Factors and Sociotechnical Systems, University of the Sunshine Coast, Sippy Downs, Queensland, Australia
| | - Scott McLean
- Centre for Human Factors and Sociotechnical Systems, University of the Sunshine Coast, Sippy Downs, Queensland, Australia
| |
Collapse
|
2
|
Möck S, Wirth K. Bilateral differences of isokinetic knee extensor strength are velocity- and task-dependent. Sports Biomech 2024:1-13. [PMID: 38329274 DOI: 10.1080/14763141.2024.2315260] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 12/21/2023] [Accepted: 01/31/2024] [Indexed: 02/09/2024]
Abstract
The purpose of this study was to investigate the concordance of isokinetic bilateral strength differences of the knee extensors in single- and multi-joint movement tasks. One hundred and nineteen male athletes performed isokinetic legpresses at 0.1 m/s and 0.7 m/s as well as isokinetic knee extensions at 60°/s and 180°/s. Bilateral differences and directed bilateral differences (sign indicating the direction of the difference) were calculated for all measurements. Bland-Altman-Plots were plotted to investigate if the different conditions detect bilateral differences of the same magnitude. Additionally, concordance correlations for the directed bilateral differences of the different tests were calculated to investigate magnitude and direction. The results indicate poor to fair concordance between the bilateral differences in the legpress conditions as well as between single- and multi-joint tasks. The single-joint knee extensions displayed a moderate level of agreement. Bilateral strength differences in isokinetic movement tasks are dependent on movement velocity and the nature of the task (single- or multi-joint movement) in the lower extremities. Both the value and the direction of the strength differences show no clear pattern across the investigated measurements and cannot be used interchangeably. Therefore, to assess interlimb strength balance, multiple different tests should be performed.
Collapse
Affiliation(s)
- Sebastian Möck
- Department of Exercise Science, Olympic Training and Testing Center of Hessen, Frankfurt am Main, Germany
| | - Klaus Wirth
- Sport and Exercise Sciences, University of Applied Sciences Wiener Neustadt, Wiener Neustadt, Austria
| |
Collapse
|
3
|
Amendolara A, Pfister D, Settelmayer M, Shah M, Wu V, Donnelly S, Johnston B, Peterson R, Sant D, Kriak J, Bills K. An Overview of Machine Learning Applications in Sports Injury Prediction. Cureus 2023; 15:e46170. [PMID: 37905265 PMCID: PMC10613321 DOI: 10.7759/cureus.46170] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Accepted: 09/28/2023] [Indexed: 11/02/2023] Open
Abstract
Use injuries, i.e., injuries caused by repetitive strain on the body, represent a serious problem in athletics that has traditionally relied on historic datasets and human experience for prevention. Existing methodologies have been frustratingly slow at developing higher precision prevention practices. Technological advancements have permitted the emergence of artificial intelligence and machine learning (ML) as promising toolsets to enhance both injury mitigation and rehabilitation protocols. This article provides a comprehensive overview of recent advances in ML techniques as they have been applied to sports injury prediction and prevention. A comprehensive literature review was conducted searching PubMed/Medline, Institute of Electrical and Electronics Engineers (IEEE)/Institute of Engineering and Technology (IET), and ScienceDirect. Ovid Discovery and Google Scholar were used to provide additional aggregate results and a grey literature search. A focus was placed on papers published from 2017 to 2022. Algorithms of interest were limited to K-Nearest Neighbor (KNN), K-means, decision tree, random forest, gradient boosting and AdaBoost, and neural networks. A total of 42 original research papers were included, and their results were summarized. We conclude that given the current lack of open source, uniform data sets, as well as a reliance on dated regression models, no strong conclusions about the real-world efficacy of ML as it applies to sports injury prediction can be made. However, it is suggested that addressing these two issues will allow powerful, novel ML architectures to be deployed, thus rapidly advancing the state of this field, and providing validated clinical tools.
Collapse
Affiliation(s)
- Alfred Amendolara
- Federated Department of Biology, New Jersey Institute of Technology, Newark, USA
- Department of Biomedical Sciences, Noorda College of Osteopathic Medicine, Provo, USA
| | - Devin Pfister
- Department of Biomedical Sciences, Noorda College of Osteopathic Medicine, Provo, USA
| | - Marina Settelmayer
- Department of Biomedical Sciences, Noorda College of Osteopathic Medicine, Provo, USA
| | - Mujtaba Shah
- Department of Biomedical Sciences, Noorda College of Osteopathic Medicine, Provo, USA
| | - Veronica Wu
- Department of Biomedical Sciences, Noorda College of Osteopathic Medicine, Provo, USA
| | - Sean Donnelly
- Department of Biomedical Sciences, Noorda College of Osteopathic Medicine, Provo, USA
| | - Brooke Johnston
- Department of Biomedical Sciences, Noorda College of Osteopathic Medicine, Provo, USA
| | - Race Peterson
- Department of Biomedical Sciences, Noorda College of Osteopathic Medicine, Provo, USA
| | - David Sant
- Department of Biomedical Sciences, Noorda College of Osteopathic Medicine, Provo, USA
| | - John Kriak
- Department of Biomedical Sciences, Noorda College of Osteopathic Medicine, Provo, USA
| | - Kyle Bills
- Department of Biomedical Sciences, Noorda College of Osteopathic Medicine, Provo, USA
| |
Collapse
|
4
|
Dainese P, Booysen N, Mulasso A, Roppolo M, Stokes M. Movement retraining programme in young soccer and rugby football players: A feasibility and proof of concept study. J Bodyw Mov Ther 2023; 33:28-38. [PMID: 36775523 DOI: 10.1016/j.jbmt.2022.09.017] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/24/2021] [Revised: 05/30/2022] [Accepted: 09/18/2022] [Indexed: 10/14/2022]
Abstract
INTRODUCTION Movement screening to identify abnormal movement patterns can inform development of effective interventions. The primary objective of this study was to evaluate the feasibility of using a movement screening tool in combination with a tailored movement control retraining programme in young soccer and rugby football players. A secondary objective was to investigate changes in movement control patterns post-intervention, to provide proof of concept (PoC) for movement retraining. METHODS 52 male amateur players, including 34 soccer players (mean age 15 ± 2 years) and 18 rugby players (mean age 15 ± 1 years) participated. They were screened for movement control ability using a shortened version of the Hip and Lower Limb Movement Screening (Short-HLLMS) and completed an eight-week movement control retraining programme. Evaluation of feasibility included consent from players invited, adherence, attendance at the exercise sessions, drop-out and adverse events. Short-HLLMS total score and The Copenhagen Hip and Groin Outcome Score (HAGOS) were analysed to provide PoC for retraining movement control. RESULTS feasibility outcomes were favourable. Significant statistical changes occurred post-intervention in the Short-HLLMS total score (paired-samples t-test) and in three HAGOS subscales (symptoms, physical function in daily living and in sport and recreation) (Wilcoxon-Signed Rank Test) in both groups. CONCLUSIONS Feasibility of using the Short-HLLMS in combination with a movement control retraining programme in soccer and rugby players was promising. The data provided PoC for the potential application of a shortened version of the HLLMS to evaluate changes in movement control and to inform targeted motor control programmes.
Collapse
Affiliation(s)
- Paolo Dainese
- Department of Rehabilitation Sciences and Physiotherapy, Ghent University, Gent, Belgium; School of Exercise and Sport Science, University of Torino, Torino, Italy.
| | - Nadine Booysen
- School of Health Sciences, University of Southampton, Southampton, UK; Centre for Sport, Exercise and Osteoarthritis Research Versus Arthritis, UK
| | - Anna Mulasso
- NeuroMuscular Function
- Research Group, School of Exercise and Sport Sciences, Department of Medical Sciences, University of Torino, Torino, Italy
| | | | - Maria Stokes
- School of Health Sciences, University of Southampton, Southampton, UK; Centre for Sport, Exercise and Osteoarthritis Research Versus Arthritis, UK; Southampton National Institute for Health Research Biomedical Research Centre, UK
| |
Collapse
|
5
|
Hecksteden A, Schmartz GP, Egyptien Y, Aus der Fünten K, Keller A, Meyer T. Forecasting football injuries by combining screening, monitoring and machine learning. SCI MED FOOTBALL 2022:1-15. [PMID: 35757889 DOI: 10.1080/24733938.2022.2095006] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 10/17/2022]
Abstract
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.
Collapse
Affiliation(s)
- Anne Hecksteden
- Saarland University, Institute of Sports and Preventive Medicine, Saarbruecken, Germany
| | | | - Yanni Egyptien
- Saarland University, Institute of Sports and Preventive Medicine, Saarbruecken, Germany
| | - Karen Aus der Fünten
- Saarland University, Institute of Sports and Preventive Medicine, Saarbruecken, Germany
| | - Andreas Keller
- Saarland University, Chair for Clinical Bioinformatics, Saarbruecken, Germany
| | - Tim Meyer
- Saarland University, Institute of Sports and Preventive Medicine, Saarbruecken, Germany
| |
Collapse
|