Speakers
Description
Objectives: Recent technological innovations have resulted in machine learning methodologies being tested to improve injury prevention, diagnosis, and treatment of sports-related concussion (SRC). The aim of this study is to (1) map existing research on the use of machine learning in the management of SRC and (2) ascertain where there are knowledge gaps and identify recommendations for future research.
Methods: Scoping review of three databases: SportsDiscus, PubMed, and Scopus, using search and mesh terms related to “machine learning”, “artificial intelligence” AND “concussion” or “head injury” or “mild traumatic brain injury” AND “sports” or “athletes”.
Findings: 24 studies were identified and analysed. Our review shows that current research literature primarily; (a) focus on predicting and preventing SRC (n= 10), diagnosing and classifying SRC (n= 11), while few explore treatment and rehabilitation of SRC (n= 3), (b) employ machine learning algorithms to analyse athlete injury data derived from male team sport athletes in North America.
Discussion: The research field on the use of machine learning in the management of SRC is novel and the potential is still unclear. Based on our review we recommend future studies to; (a) use more diverse datasets and samples, (b) explore the potential of deep learning, and (c) include social scientific perspectives on the implementation of novel technological tools created through machine learning approaches.
Conclusion: Studies using more diverse athlete data for analysis through machine learning algorithms have the potential to fill important knowledge gaps and improve athletes’ welfare by reducing the risk of sustaining SRC, as well as providing better diagnosis and treatment.
