No JavaScript support!

To take full advantage of this digital library, please verify if your browser supports JavaScript.

polski

English

dLibra Digital Library

readers on-line: 457

Edition

Export metadata

Favourite positions

Add to Favourites

Add to Favourites

Please login or register to use additional features!

Save this address

Add to bookmarks

Edition description

Artificial Intelligence in Injury Prediction: A Review of Ethical Practices and Technical Standards in Elite Football

Show publication content!

Show publication content!

Artificial Intelligence in Injury Prediction: A Review of Ethical Practices and Technical Standards in Elite Football

Waśkiewicz, Zbigniew ; Grzywacz, Tomasz ; Słomka, Kajetan ; Jurkojć, Jacek ; Juras, Grzegorz

artificial intelligence ; injury prediction ; professional soccer ; algorithmic bias ; explainable AI ; athlete autonomy ; ethics in sports technology

Physical Activity Review ; Vol. 14 ; Issue 1 ; pp. 26-41

The use of artificial intelligence (AI) for injury prediction in professional soccer has expanded rapidly, offering promising tools for performance optimization and medical decision-making. However, integrating AI into athlete health monitoring raises critical ethical concerns, particularly regarding algorithmic bias, model transparency, and player autonomy. This systematic review assesses the existing literature on AI-based injury prediction systems in professional soccer, with a focus on ethical dimensions, including bias mitigation, explainability, stakeholder rights, and governance frameworks. A systematic search was conducted in January 2025 across the Semantic Scholar database, yielding 497 records. Following PRISMA 2020 guidelines, 14 studies were included in the final synthesis after duplicate removal and full-text eligibility assessment. Extracted data focused on AI model types, bias identification, mitigation strategies, ethical safeguards, and stakeholder considerations. The included studies employed various AI techniques, including decision trees, XGBoost, support vector machines, and neural networks. While performance metrics were frequently reported, only six studies addressed ethical concerns. Common challenges included class imbalance, limited generalizability, and lack of transparency. Few studies discussed data ownership, consent, or the downstream implications of algorithmic decisions on player welfare. The current literature demonstrates a growing technical focus on AI-driven injury prediction but lacks robust integration of ethical and bias-related considerations. Future development of such systems should prioritize fairness audits, athlete-centered governance, and explainable AI to ensure responsible and equitable application in elite sports contexts.

Place of publishing:

Częstochowa

Projack

2026

artykuł

application/pdf

e-ISSN 2300-5076 ; oai:dlibra.bg.ajd.czest.pl:8346

10.16926/par.2026.14.03 ; click here to follow the link

eng

click here to follow the link

Uniwersytet Jana Długosza w Częstochowie

Location of original object:

Biblioteka Uniwersytecka Uniwersytetu Jana Długosza w Częstochowie