A Machine Learning Approach to Concussion Risk Estimation Among Players Exhibiting Visible Signs in Professional Hockey.


Journal

Sports medicine (Auckland, N.Z.)
ISSN: 1179-2035
Titre abrégé: Sports Med
Pays: New Zealand
ID NLM: 8412297

Informations de publication

Date de publication:
17 Sep 2024
Historique:
accepted: 21 08 2024
medline: 17 9 2024
pubmed: 17 9 2024
entrez: 17 9 2024
Statut: aheadofprint

Résumé

The identification of concussion risk factors, such as visible signs and mechanisms of injury, improves concussion identification. Exploring individual risk factors, such as concussion history, may help to improve existing concussion risk models and algorithms. The primary aim of the current study was to use machine learning techniques to develop a comprehensive, prospectively coded concussion risk model in professional hockey among players exhibiting visible signs. The secondary aim was to examine whether including concussion history improves model performance. Data from the National Hockey League (NHL) spotter program, including coded visible signs and mechanisms of injury associated with possible concussive events, were extracted from the 2018-2019 to the 2021-2022 seasons. Each unique spotter event was matched with data extracted from the medical record to determine whether the event was associated with a subsequent physician diagnosed concussion. We compared the ability of three machine learning-based approaches to identify the likelihood of physician diagnosed concussion: conditional inference tree, conditional inference random forest, and logistic regression. A total of 1563 unique events with visible signs were identified by spotters (183 leading to a concussion diagnosis). A randomly selected training sample had 1250 events (146 concussions) and the remaining set-aside test sample had 313 events (37 concussions). The obtained models performed at a high level with large effects in the training [area under the receiver operating characteristic curve (AUC) = 0.79] and set-aside test data (AUC = 0.82). Concussion history was retained in the tree and logistic regression models, with each additional prior concussion associated with a 1.32 times increased odds of concussion diagnosis. We present simple tree and logistic algorithms for concussion screening and as diagnostic aids. Our results show that player concussion history can explain additional risk above and beyond that explained by visible signs and mechanisms of injury alone.

Sections du résumé

BACKGROUND BACKGROUND
The identification of concussion risk factors, such as visible signs and mechanisms of injury, improves concussion identification. Exploring individual risk factors, such as concussion history, may help to improve existing concussion risk models and algorithms.
OBJECTIVES OBJECTIVE
The primary aim of the current study was to use machine learning techniques to develop a comprehensive, prospectively coded concussion risk model in professional hockey among players exhibiting visible signs. The secondary aim was to examine whether including concussion history improves model performance.
METHODS METHODS
Data from the National Hockey League (NHL) spotter program, including coded visible signs and mechanisms of injury associated with possible concussive events, were extracted from the 2018-2019 to the 2021-2022 seasons. Each unique spotter event was matched with data extracted from the medical record to determine whether the event was associated with a subsequent physician diagnosed concussion. We compared the ability of three machine learning-based approaches to identify the likelihood of physician diagnosed concussion: conditional inference tree, conditional inference random forest, and logistic regression.
RESULTS RESULTS
A total of 1563 unique events with visible signs were identified by spotters (183 leading to a concussion diagnosis). A randomly selected training sample had 1250 events (146 concussions) and the remaining set-aside test sample had 313 events (37 concussions). The obtained models performed at a high level with large effects in the training [area under the receiver operating characteristic curve (AUC) = 0.79] and set-aside test data (AUC = 0.82). Concussion history was retained in the tree and logistic regression models, with each additional prior concussion associated with a 1.32 times increased odds of concussion diagnosis.
CONCLUSIONS CONCLUSIONS
We present simple tree and logistic algorithms for concussion screening and as diagnostic aids. Our results show that player concussion history can explain additional risk above and beyond that explained by visible signs and mechanisms of injury alone.

Identifiants

pubmed: 39287776
doi: 10.1007/s40279-024-02112-2
pii: 10.1007/s40279-024-02112-2
doi:

Types de publication

Journal Article

Langues

eng

Sous-ensembles de citation

IM

Informations de copyright

© 2024. The Author(s), under exclusive licence to Springer Nature Switzerland AG.

Références

Asken BM, McCrea MA, Clugston JR, et al. “Playing through it”: delayed reporting and removal from athletic activity after concussion predicts prolonged recovery. J Athl Train. 2016;51(5):329–35.
pubmed: 27111584 pmcid: 4874376
McCrea M, Hammeke T, Olsen G, et al. Unreported concussion in high school football players. Clin J Sport Med. 2004;14(1):13–7. https://doi.org/10.1097/00042752-200401000-00003 .
pubmed: 14712161
Schmidt JD, Broglio SP, Knight K, et al. Optimizing concussion care seeking: a longitudinal analysis of recovery. Am J Sports Med. 2022. https://doi.org/10.1177/03635465221135771 .
pubmed: 36412549
Patricios JS, Schneider KJ, Dvorak J, et al. Consensus statement on concussion in sport: the 6th international conference on concussion in Sport-Amsterdam, October 2022. Br J Sports Med. 2023;57(11):695–711. https://doi.org/10.1136/bjsports-2023-106898 .
pubmed: 37316210
Echemendia RJ, Burma JS, Bruce, et al. Acute evaluation of sport-related concussion and implications for the Sport Concussion Assessment Tool (SCAT6) for adults, adolescents and children: a systematic review. Br J Sports Med. 2023;57(11):722–35. https://doi.org/10.1136/bjsports-2022-106661 .
pubmed: 37316213
Bruce JM, Echemendia RJ, Meeuwisse W, et al. Development of a risk prediction model among professional hockey players with visible signs of concussion. Br J Sports Med. 2018;52(17):1143–8. https://doi.org/10.1136/bjsports-2016-097091 .
pubmed: 28377444
Echemendia RJ, Bruce JM, Meeuwisse W, et al. Can visible signs predict concussion diagnosis in the National Hockey League? Br J Sports Med. 2018;52(17):1149–54. https://doi.org/10.1136/bjsports-2016-097090 .
pubmed: 28377445
Gardner AJ, Howell DR, Iverson GL. A video review of multiple concussion signs in National Rugby League match play. Sports Med Open. 2018;4(1):0–7. https://doi.org/10.1186/s40798-017-0117-9 .
Gardner AJ, Howell DR, Levi CR, Iverson GL. Evidence of concussion signs in national rugby league match play: a video review and validation study. Sports Med Open. 2017. https://doi.org/10.1186/s40798-017-0097-9 .
pubmed: 28831759 pmcid: 5567587
Makdissi M, Davis G. The reliability and validity of video analysis for the assessment of the clinical signs of concussion in Australian football. J Sci Med Sport. 2016;19(10):859–63. https://doi.org/10.1016/j.jsams.2016.02.015 .
pubmed: 27009774
Abrahams S, McFie S, Patricios J, et al. Risk factors for sports concussion: an evidence-based systematic review. Br J Sports Med. 2014;48(2):91–7. https://doi.org/10.1136/bjsports-2013-092734 .
pubmed: 24052371
Putukian M, Riegler K, Amalfe S, et al. Preinjury and postinjury factors that predict sports-related concussion and clinical recovery time. Clin J Sport Med. 2018;00(00):1. https://doi.org/10.1097/JSM.0000000000000705 .
Bergeron MF, Landset S, Maugans, et al. Machine learning in modeling high school sport concussion symptom resolve. Med Sci Sports Exerc. 2019;51(7):1362–71. https://doi.org/10.1249/MSS.0000000000001903 .
pubmed: 30694980
Bittencourt NFN, Meeuwisse WH, Mendonça LD, et al. Complex systems approach for sports injuries: moving from risk factor identification to injury pattern recognition—narrative review and new concept. Br J Sports Med. 2016;50(21):1309–14. https://doi.org/10.1136/bjsports-2015-095850 .
pubmed: 27445362
Kampakis S. Comparison of machine learning methods for predicting the recovery time of professional football players after an undiagnosed injury. CEUR Workshop Proc. 2013;1969:58–68.
Blatt SJ, Luyten P. A structural-developmental psychodynamic approach to psychopathology: Two polarities of experience across the life span. Dev Psychopathol. 2009;21(3):793–814. https://doi.org/10.1017/S0954579409000431 .
pubmed: 19583884
Rstudio Team. Rstudio: integrated development environment for R. RStudio, PBC, Boston, MA; 2022. http://www.rstudio.com/ . Accessed 12 Dec 2022
Jiang T, Gradus JL, Rosellini AJ. Supervised machine learning: a brief primer. Behav Ther. 2020;51(5):675–87. https://doi.org/10.1016/j.beth.2020.05.002 .
pubmed: 32800297 pmcid: 7431677
Kuhm M, Johnson K. Applied predictive modeling. Berlin: Springer; 2018.
Sidey-Gibbons J, Sidey-Gibbons C. Machine learning in medicine: a practical introduction. BMC Med Res Methodol. 2019;19:64. https://doi.org/10.1186/s12874-019-0681-4 .
pubmed: 30890124 pmcid: 6425557
Bruce JM, Thelen J, Meeuwisse W, et al. Use of the sport concussion assessment tool 5 (SCAT5) in professional hockey, part 2: which components differentiate concussed and non-concussed players? Br J Sports Med. 2021;55(10):557–65. https://doi.org/10.1136/bjsports-2020-102072 .
Torres DM, Galetta KM, Phillips HW, et al. Sports-related concussion: anonymous survey of a collegiate cohort. Neurol Clin Pract. 2013;3(4):279–87.
pubmed: 24195017 pmcid: 3787116
Kroshus E, Chrisman SP, Milroy JJ, Baugh CM. History of concussion diagnosis, differences in concussion reporting behavior, and self-described reasons for non-report. J Clin Sport Psychol. 2020;14(1):41–54. https://doi.org/10.1123/jcsp.2017-0036 .
Craig DI, Lininger MR, Wayment HA, Huffman AH. Investigation of strategies to improve concussion reporting in American football. Res Sports Med. 2020;28(2):181–93. https://doi.org/10.1080/15438627.2019.1586706 .
pubmed: 30859851
Rahnama N, Reilly T, Lees A. Injury risk associated with playing actions during competitive soccer. Br J Sports Med. 2002;36(5):354–9. https://doi.org/10.1136/bjsm.36.5.354 .
pubmed: 12351333 pmcid: 1724551
O’Connor KL, Rowson S, Duma SM, Broglio SP. Head-impact-measurement devices: a systematic review. J Athl Train. 2017;52(3):206–27. https://doi.org/10.4085/1062-6050.52 .
pubmed: 28387553 pmcid: 5384819

Auteurs

Jared M Bruce (JM)

Department of Biomedical and Health Informatics, University of Missouri-Kansas City School of Medicine, Kansas City, MO, 64108, USA. brucejm@umkc.edu.
Department of Neurology, University Health, Kansas City, MO, 64108, USA. brucejm@umkc.edu.
Department of Psychiatry, University Health, Kansas City, MO, 64108, USA. brucejm@umkc.edu.

Kaitlin E Riegler (KE)

Princeton Neuropsychology and Sports Concussion Center of New Jersey at RSM Psychology, Princeton, NJ, 08540, USA.

Willem Meeuwisse (W)

The National Hockey League, New York, NY, USA.

Paul Comper (P)

Faculty of Kinesiology and Physical Education, University of Toronto, Toronto, ON, MS55 1A1, Canada.
Toronto Rehabilitation Institute, Toronto, ON, M5G 2A2, Canada.

Michael G Hutchison (MG)

Faculty of Kinesiology and Physical Education, University of Toronto, Toronto, ON, MS55 1A1, Canada.

J Scott Delaney (JS)

McGill Sport Medicine Clinic, Montreal, QC, Canada.
Department of Emergency Medicine, McGill University Health Centre, Montreal, QC, Canada.

Ruben J Echemendia (RJ)

Psychological and Neurobehavioral Associates, Inc., State College, PA, 16801, USA.

Classifications MeSH