Previsão de lesões de esportistas utilizando técnicas de aprendizagem de máquina em uma revisão de literatura

Abstract

Sports injuries pose a significant concern within collective sports, leading to a growing demand for tools that aid in predicting such occurrences. This article presents a literature review focused on the application of Machine Learning in predicting injuries among athletes in collective sports, encompassing studies published between 2018 and 2025. A total of 21 articles were analyzed, revealing diversity in the Machine Learning algorithms employed, including Random Forest, Decision Tree, Neural Networks, and Time Series Analysis. The main factors considered in the predictions include training load, individual athlete characteristics, and psychological aspects. Results indicate that, despite the promising potential of these approaches, limitations persist regarding the quality of the data used in model development and the lack of external validation to confirm prediction efficacy in practical contexts. The study also explores the application of similar models in injury rehabilitation through movement monitoring. Finally, it is recommended that data collection methods be standardized and that the interpretability of predictions be enhanced through Explainable AI techniques.