{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T13:31:03Z","timestamp":1787923863271,"version":"build-2784847793"},"reference-count":35,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T00:00:00Z","timestamp":1787875200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Public Health"],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>This study evaluates whether seven-day training-load histories can support athlete-independent classification of injury-labeled days in competitive distance runners. The aim is methodological benchmarking rather than clinical injury-risk deployment.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Research gap<\/jats:title>\n                    <jats:p>Prior sports-injury prediction studies often use repeated athlete-day observations without fully addressing athlete leakage, rare-event imbalance, probability calibration, and the risk that explainability methods may identify recording artifacts rather than physiological mechanisms. Data and method: The public dataset contained 42,766 athlete-day observations from 74 runners, including 583 injury-labeled observations (1.36%). Seven external-load variables and three internal-response variables were organized from Day \u22127 to Day \u22121. A compact dual-stream multi-scale temporal convolutional network with gated fusion and feature-by-day Integrated Gradients was evaluated using athlete-independent five-fold grouped cross-validation. Logistic regression and XGBoost served as baselines, while three ablations and five prespecified neural-network seeds assessed robustness.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Findings<\/jats:title>\n                    <jats:p>\n                      In the primary analysis, DS-MTCN-IG achieved ROC-AUC 0.596, PR-AUC 0.0204, and Brier score 0.01343. Logistic regression achieved a numerically higher ROC-AUC of 0.619, but the paired athlete-cluster bootstrap difference was not statistically significant (\u0394ROC-AUC\u202f=\u202f0.0227, 95% CI\u202f\u2212\u202f0.0180 to 0.0714;\n                      <jats:italic>p<\/jats:italic>\n                      \u202f=\u202f0.291). Across five seeds, the ungated dual-stream model achieved the highest mean ROC-AUC among the deep variants (0.604 vs. 0.598 for the gated model), so the gating hypothesis was not supported. In the non-overlapping evaluation, which removed temporal adjacency between successive prediction targets, DS-MTCN-IG achieved a mean ROC-AUC of 0.469 and did not demonstrate above-chance discrimination. This finding indicates that the weak discrimination observed in the full-window analysis may reflect temporal overlap and autocorrelation between adjacent seven-day histories rather than a stable athlete-independent relationship between training-load history and injury-labeled days. Five-seed Integrated Gradients identified contributions from measured internal-response ratings and availability masks, but exact top-input rankings were unstable across seeds.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>The study did not demonstrate a robust or practically useful athlete-independent predictive signal. The non-overlapping null result substantially limits interpretation of the full-window estimates and provides no support for clinical, coaching, or automated injury-warning use. The principal contribution is therefore a methodological audit showing how temporal overlap, participant-aware validation, calibration, model comparison, and attribution stability can alter conclusions in longitudinal sports-injury modeling.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fpubh.2026.1936365","type":"journal-article","created":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T12:51:08Z","timestamp":1787921468000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Explainable temporal deep learning for athlete-independent classification of injury-labeled days in competitive distance runners: a methodological benchmark using seven-day training-load histories"],"prefix":"10.3389","volume":"14","author":[{"given":"Tianyi","family":"Qiao","sequence":"first","affiliation":[{"name":"Faculty of Sports, Belarusian National Technical University","place":["Minsk, Belarus"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Wenjun","family":"Tian","sequence":"additional","affiliation":[{"name":"College of Music, Ewha Womans University","place":["Seoul, Republic of Korea"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,8,28]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"513","DOI":"10.1016\/j.jshs.2021.04.001","article-title":"A systematic review of running-related musculoskeletal injuries in runners","volume":"10","author":"Kakouris","year":"2021","journal-title":"J Sport Health Sci"},{"key":"ref2","doi-asserted-by":"publisher","first-page":"38","DOI":"10.1186\/s40798-022-00416-z","article-title":"Biomechanical and musculoskeletal measurements as risk factors for running-related injury in non-elite runners: a systematic review and meta-analysis of prospective studies","volume":"8","author":"Peterson","year":"2022","journal-title":"Sports Med Open"},{"key":"ref3","doi-asserted-by":"publisher","first-page":"1863","DOI":"10.1007\/s40279-022-01666-3","article-title":"Running-related biomechanical risk factors for overuse injuries in distance runners: a systematic review considering injury specificity and the potentials for future research","volume":"52","author":"Willwacher","year":"2022","journal-title":"Sports Med"},{"key":"ref4","doi-asserted-by":"publisher","first-page":"1170","DOI":"10.1111\/sms.12883","article-title":"A framework for the etiology of running-related injuries","volume":"27","author":"Bertelsen","year":"2017","journal-title":"Scand J Med Sci Sports"},{"key":"ref5","doi-asserted-by":"publisher","first-page":"665683","DOI":"10.3389\/fspor.2021.665683","article-title":"Training load capacity, cumulative risk, and bone stress injuries: a narrative review of a holistic approach","volume":"3","author":"Hamstra-Wright","year":"2021","journal-title":"Front Sports Active Living"},{"key":"ref6","doi-asserted-by":"publisher","first-page":"611550","DOI":"10.3389\/fphys.2020.611550","article-title":"Network physiology of exercise: vision and perspectives","volume":"11","author":"Balagu\u00e9","year":"2020","journal-title":"Front Physiol"},{"key":"ref7","doi-asserted-by":"publisher","first-page":"630576","DOI":"10.3389\/fspor.2020.630576","article-title":"Wearable technology and analytics as a complementary toolkit to optimize workload and to reduce injury burden","volume":"2","author":"Seshadri","year":"2021","journal-title":"Front Sports Active Living"},{"key":"ref8","doi-asserted-by":"publisher","first-page":"241","DOI":"10.1007\/s40279-022-01760-6","article-title":"Wearables for running gait analysis: a systematic review","volume":"53","author":"Stuart","year":"2023","journal-title":"Sports Med"},{"key":"ref9","doi-asserted-by":"publisher","first-page":"491","DOI":"10.1136\/bjsports-2024-108576","article-title":"Machine learning approaches to injury risk prediction in sport: a scoping review with evidence synthesis","volume":"59","author":"Leckey","year":"2025","journal-title":"Br J Sports Med"},{"key":"ref10","doi-asserted-by":"publisher","first-page":"2516","DOI":"10.3390\/diagnostics14222516","article-title":"Diagnostic applications of AI in sports: a comprehensive review of injury risk prediction methods","volume":"14","author":"Mu\u0219at","year":"2024","journal-title":"Diagnostics"},{"key":"ref11","doi-asserted-by":"publisher","first-page":"27","DOI":"10.1186\/s40634-021-00346-x","article-title":"Machine learning methods in sport injury prediction and prevention: a systematic review","volume":"8","author":"Van Eetvelde","year":"2021","journal-title":"J Experimental Orthopaedics"},{"key":"ref12","doi-asserted-by":"publisher","first-page":"1522","DOI":"10.1123\/ijspp.2020-0518","article-title":"Injury prediction in competitive runners with machine learning","volume":"16","author":"L\u00f6vdal","year":"2021","journal-title":"Int J Sports Physiol Perform"},{"key":"ref13","first-page":"1522","article-title":"Injury prediction in competitive runners with machine learning","volume-title":"Int J Sports Physiol Perform","author":"L\u00f6vdal","year":"2021"},{"key":"ref14","doi-asserted-by":"publisher","first-page":"1174525","DOI":"10.3389\/fphys.2023.1174525","article-title":"A novel approach for sports injury risk prediction: based on time-series image encoding and deep learning","volume":"14","author":"Ye","year":"2023","journal-title":"Front Physiol"},{"key":"ref15","doi-asserted-by":"publisher","first-page":"937546","DOI":"10.3389\/fphys.2022.937546","article-title":"A novel lower extremity non-contact injury risk prediction model based on multimodal fusion and interpretable machine learning","volume":"13","author":"Huang","year":"2022","journal-title":"Front Physiol"},{"key":"ref16","doi-asserted-by":"publisher","first-page":"511","DOI":"10.1080\/17461391.2021.1887369","article-title":"Personalized machine learning approach to injury monitoring in elite volleyball players","volume":"22","author":"de Leeuw","year":"2022","journal-title":"Eur J Sport Sci"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"e0118432","DOI":"10.1371\/journal.pone.0118432","article-title":"The precision-recall plot is more informative than the ROC plot when evaluating binary classifiers on imbalanced datasets","volume":"10","author":"Saito","year":"2015","journal-title":"PLoS One"},{"key":"ref18","doi-asserted-by":"publisher","first-page":"6","DOI":"10.1186\/s12864-019-6413-7","article-title":"The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation","volume":"21","author":"Chicco","year":"2020","journal-title":"BMC Genomics"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"230","DOI":"10.1186\/s12916-019-1466-7","article-title":"Calibration: the Achilles heel of predictive analytics","volume":"17","author":"Van Calster","year":"2019","journal-title":"BMC Med"},{"key":"ref20","doi-asserted-by":"publisher","first-page":"247","DOI":"10.1109\/JPROC.2021.3060483","article-title":"Explaining deep neural networks and beyond: a review of methods and applications","volume":"109","author":"Samek","year":"2021","journal-title":"Proc IEEE"},{"key":"ref21","doi-asserted-by":"publisher","first-page":"4793","DOI":"10.1109\/TNNLS.2020.3027314","article-title":"A survey on explainable artificial intelligence (XAI): toward medical XAI","volume":"32","author":"Tjoa","year":"2021","journal-title":"IEEE Trans Neural Networks Learn Syst"},{"key":"ref22","doi-asserted-by":"publisher","first-page":"84","DOI":"10.1186\/s40798-024-00745-1","article-title":"Comment on: machine learning for understanding and predicting injuries in football","volume":"10","author":"Bullock","year":"2024","journal-title":"Sports Med Open"},{"key":"ref23","volume-title":"National Fitness Program (2021\u20132025)","year":"2021"},{"key":"ref24","doi-asserted-by":"publisher","first-page":"581","DOI":"10.1007\/s40279-020-01378-6","article-title":"What role do chronic workloads play in the acute to chronic workload ratio? Time to dismiss ACWR and its underlying theory","volume":"51","author":"Impellizzeri","year":"2021","journal-title":"Sports Med"},{"key":"ref25","doi-asserted-by":"publisher","first-page":"216","DOI":"10.1123\/ijspp.2021-0190","article-title":"Effects of increased load of low- versus high-intensity endurance training on performance and physiological adaptations in endurance athletes","volume":"17","author":"Talsnes","year":"2022","journal-title":"Int J Sports Physiol Perform"},{"key":"ref26","doi-asserted-by":"publisher","first-page":"73","DOI":"10.1186\/s40798-022-00465-4","article-title":"Machine learning for understanding and predicting injuries in football","volume":"8","author":"Majumdar","year":"2022","journal-title":"Sports Med Open"},{"key":"ref27","doi-asserted-by":"publisher","first-page":"1745","DOI":"10.1249\/MSS.0000000000002305","article-title":"A machine learning approach to assess injury risk in elite youth football players","volume":"52","author":"Rommers","year":"2020","journal-title":"Med Sci Sports Exerc"},{"key":"ref28","author":"Murugan","year":"2025"},{"key":"ref29","author":"Kim","year":"2019"},{"key":"ref30","author":"Sundararajan","year":"2017"},{"key":"ref31","author":"Guo","year":"2017"},{"key":"ref32","doi-asserted-by":"publisher","first-page":"e078378","DOI":"10.1136\/bmj-2023-078378","article-title":"TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods","volume":"385","author":"Collins","year":"2024","journal-title":"BMJ"},{"key":"ref33","doi-asserted-by":"publisher","first-page":"e082505","DOI":"10.1136\/bmj-2024-082505","article-title":"PROBAST+AI: an updated quality, risk of bias, and applicability assessment tool for prediction models using regression or artificial intelligence methods","volume":"388","author":"Moons","year":"2025","journal-title":"BMJ"},{"key":"ref34","author":"Lin","year":"2017"},{"key":"ref35","author":"Loshchilov","year":"2019"}],"container-title":["Frontiers in Public Health"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fpubh.2026.1936365\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T12:51:15Z","timestamp":1787921475000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fpubh.2026.1936365\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,28]]},"references-count":35,"alternative-id":["10.3389\/fpubh.2026.1936365"],"URL":"https:\/\/doi.org\/10.3389\/fpubh.2026.1936365","relation":{},"ISSN":["2296-2565"],"issn-type":[{"value":"2296-2565","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,28]]},"article-number":"1936365"}}