{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,22]],"date-time":"2026-09-22T16:17:47Z","timestamp":1790093867468,"version":"4.0.1"},"reference-count":44,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,1,12]],"date-time":"2026-01-12T00:00:00Z","timestamp":1768176000000},"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. Neurol."],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Stroke is one of the leading causes of death and disability worldwide, making early screening and risk prediction crucial. Traditional methods have limitations in handling nonlinear relationships between variables, class imbalance, and model interpretability.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>Logistic regression (LR), random forest (RF), extreme gradient boosting (XGBoost), categorical boosting (CatBoost), multi-layer perceptron (MLP) neural network, and ensemble models were constructed and compared. Their performance in stroke risk prediction was systematically evaluated, and feature contributions were interpreted using SHapley Additive exPlanations (SHAP). Confusion matrices and Precision-Recall (PR) curves were used to compare the differences in recognition of the positive class (stroke patients) among the models, and training time was calculated to quantify resource consumption.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>The ensemble model and neural network demonstrated superior overall predictive ability to traditional algorithms, with the MLP performing particularly well in terms of recall. SHAP results revealed that \u201chypertension,\u201d \u201caverage blood glucose level,\u201d and \u201cage\u201d were key influencing factors. Confusion matrices and PR curves indicated differences in positive classification among the models. Training time analysis provided a basis for resource assessment for subsequent deployment.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>Machine learning methods have advantages in stroke risk prediction. Incorporating interpretability analysis can enhance the clinical credibility of the models, providing data and methodological reference for stroke risk stratification management and early warning.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fneur.2025.1716984","type":"journal-article","created":{"date-parts":[[2026,1,12]],"date-time":"2026-01-12T05:45:55Z","timestamp":1768196755000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":9,"title":["Explainable machine learning for stroke risk prediction: a comparative study with SHAP-based interpretation"],"prefix":"10.3389","volume":"16","author":[{"given":"Xiaoyu","family":"Tang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Min","family":"Tang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Wu","family":"Liu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shaoyang","family":"Cui","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,1,12]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"399","DOI":"10.1038\/nrn1106","article-title":"Mechanisms, challenges and opportunities in stroke","volume":"4","author":"Lo","year":"2003","journal-title":"Nat Rev Neurosci"},{"key":"ref2","doi-asserted-by":"publisher","first-page":"417","DOI":"10.1016\/S1474-4422(19)30030-4","article-title":"The global burden of stroke: persistent and disabling","volume":"18","author":"Gorelick","year":"2019","journal-title":"Lancet Neurol"},{"key":"ref3","doi-asserted-by":"publisher","first-page":"418","DOI":"10.1007\/s13346-023-01408-6","article-title":"Modulation of microglial polarization by sequential targeting surface-engineered exosomes improves therapy for ischemic stroke","volume":"14","author":"Liu","year":"2024","journal-title":"Drug Deliv Transl Res"},{"key":"ref4","doi-asserted-by":"publisher","first-page":"3754","DOI":"10.1161\/STR.0000000000000046","article-title":"Guidelines for the primary prevention of stroke","volume":"45","author":"Meschia","year":"2014","journal-title":"Stroke"},{"key":"ref5","doi-asserted-by":"publisher","first-page":"e235","DOI":"10.1016\/S2214-109X(20)30520-9","article-title":"Facing the stroke burden worldwide","volume":"9","author":"Lanas","year":"2021","journal-title":"Lancet Glob Health"},{"key":"ref6","doi-asserted-by":"publisher","first-page":"1347","DOI":"10.1056\/NEJMra1814259","article-title":"Machine learning in medicine","volume":"380","author":"Rajkomar","year":"2019","journal-title":"N Engl J Med"},{"key":"ref7","doi-asserted-by":"publisher","first-page":"1317","DOI":"10.1001\/jama.2017.18391","article-title":"Big data and machine learning in health care","volume":"319","author":"Beam","year":"2018","journal-title":"JAMA"},{"key":"ref8","doi-asserted-by":"publisher","first-page":"84486","DOI":"10.1109\/ACCESS.2022.3197671","article-title":"Explainable AI for healthcare 5.0: opportunities and challenges","volume":"10","author":"Saraswat","year":"2022","journal-title":"IEEE Access"},{"key":"ref9","doi-asserted-by":"publisher","first-page":"5","DOI":"10.1023\/A:1010933404324","article-title":"Random forests","volume":"45","author":"Breiman","year":"2001","journal-title":"Mach Learn"},{"key":"ref10","author":"Chen","year":"2016"},{"key":"ref11","author":"Kumar","year":"2021"},{"key":"ref12","doi-asserted-by":"publisher","first-page":"2783","DOI":"10.1890\/07-0539.1","article-title":"Random forests for classification in ECOLOGY","volume":"88","author":"Cutler","year":"2007","journal-title":"Ecology"},{"key":"ref13","doi-asserted-by":"publisher","first-page":"3642","DOI":"10.1161\/STROKEAHA.121.034738","article-title":"Optic tract shrinkage limits visual restoration after occipital stroke","volume":"52","author":"Fahrenthold","year":"2021","journal-title":"Stroke"},{"key":"ref14","doi-asserted-by":"publisher","first-page":"1947","DOI":"10.1021\/ci034160g","article-title":"Random Forest: a classification and regression tool for compound classification and QSAR modeling","volume":"43","author":"Svetnik","year":"2003","journal-title":"J Chem Inf Comput Sci"},{"key":"ref15","doi-asserted-by":"publisher","first-page":"1189","DOI":"10.1214\/aos\/1013203451","article-title":"Greedy function approximation: a gradient boosting machine","volume":"29","author":"Friedman","year":"2001","journal-title":"Ann Stat"},{"key":"ref16","doi-asserted-by":"publisher","first-page":"11","DOI":"10.1016\/j.eswa.2017.01.041","article-title":"Automatically generating effective search queries directly from community question-answering questions for finding related questions","volume":"77","author":"Figueroa","year":"2017","journal-title":"Expert Syst Appl"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"85","DOI":"10.1016\/j.neunet.2014.09.003","article-title":"Deep learning in neural networks: an overview","volume":"61","author":"Schmidhuber","year":"2015","journal-title":"Neural Netw"},{"key":"ref18","doi-asserted-by":"publisher","first-page":"436","DOI":"10.1038\/nature14539","article-title":"Deep learning","volume":"521","author":"LeCun","year":"2015","journal-title":"Nature"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"106112","DOI":"10.1016\/j.cmpb.2021.106112","article-title":"A novel marker detection system for people with visual impairment using the improved tiny-YOLOv3 model","volume":"205","author":"Elgendy","year":"2021","journal-title":"Comput Methods Prog Biomed"},{"key":"ref20","doi-asserted-by":"publisher","first-page":"670","DOI":"10.3390\/s22134670","article-title":"Stroke risk prediction with machine learning techniques","volume":"22","author":"Dritsas","year":"2022","journal-title":"Sensors"},{"key":"ref21","doi-asserted-by":"publisher","first-page":"1263","DOI":"10.1161\/STROKEAHA.118.024293","article-title":"Machine learning\u2013based model for prediction of outcomes in acute stroke","volume":"50","author":"Heo","year":"2019","journal-title":"Stroke"},{"key":"ref22","doi-asserted-by":"publisher","first-page":"e1003945","DOI":"10.1371\/journal.pmed.1003945","article-title":"First-trimester exposure to benzodiazepines and risk of congenital malformations in offspring: a population-based cohort study in South Korea","volume":"19","author":"Noh","year":"2022","journal-title":"PLoS Med"},{"key":"ref23","doi-asserted-by":"publisher","first-page":"103798","DOI":"10.1016\/j.jbi.2021.103798","article-title":"Impact of COVID-19 on the outreach strategy of cancer social service agencies in Singapore: a pre-post analysis with Facebook data","volume":"118","author":"Ethan Tan","year":"2021","journal-title":"J Biomed Inform"},{"key":"ref24","doi-asserted-by":"publisher","first-page":"104762","DOI":"10.1016\/j.compbiomed.2021.104762","article-title":"Changes in physiological network connectivity of body system in narcolepsy during REM sleep","volume":"136","author":"Son","year":"2021","journal-title":"Comput Biol Med"},{"key":"ref25","doi-asserted-by":"publisher","first-page":"3975","DOI":"10.1002\/sim.7848","article-title":"State-level estimation of diabetes and prediabetes prevalence: combining national and local survey data and clinical data","volume":"37","author":"Marker","year":"2018","journal-title":"Stat Med"},{"key":"ref26","doi-asserted-by":"publisher","first-page":"191","DOI":"10.1155\/2023\/7352191","article-title":"Tree-based risk factor identification and stroke level prediction in stroke cohort study","volume":"2023","author":"Li","year":"2023","journal-title":"Biomed Res Int"},{"key":"ref27","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":"ref28","author":"Ribeiro","year":"2016"},{"key":"ref29","first-page":"480","article-title":"Influence of astrocytic gap junction coupling on in silico neuronal network activity","volume-title":"Mediterranean Conference on Medical and Biological Engineering and Computing","author":"Genocchi","year":"2019"},{"key":"ref30","doi-asserted-by":"publisher","first-page":"382","DOI":"10.1016\/j.cca.2019.05.004","article-title":"The VLDL receptor plays a key role in the metabolism of postprandial remnant lipoproteins","volume":"495","author":"Nakajima","year":"2019","journal-title":"Clin Chim Acta"},{"key":"ref31","doi-asserted-by":"publisher","first-page":"e24876","DOI":"10.1016\/j.heliyon.2024.e24876","article-title":"Clinical decision support tool for breast cancer recurrence prediction using SHAP value in cooperative game theory","volume":"10","author":"Liu","year":"2024","journal-title":"Heliyon"},{"key":"ref32","doi-asserted-by":"publisher","first-page":"676","DOI":"10.1093\/bib\/bbz018","article-title":"CAFU: a galaxy framework for exploring unmapped RNA-Seq data","volume":"21","author":"Chen","year":"2020","journal-title":"Brief Bioinform"},{"key":"ref33","doi-asserted-by":"publisher","first-page":"25","DOI":"10.1161\/HYPERTENSIONAHA.111.189100","article-title":"Does blood pressure inevitably rise with age?","volume":"60","author":"Gurven","year":"2012","journal-title":"Hypertension"},{"key":"ref34","doi-asserted-by":"publisher","first-page":"24","DOI":"10.1016\/j.puhe.2019.03.018","article-title":"Did school characteristics affect the uptake of meningococcal quadrivalent vaccine in greater Manchester, United Kingdom?","volume":"171","author":"Fletcher","year":"2019","journal-title":"Public Health"},{"key":"ref35","doi-asserted-by":"publisher","first-page":"207","DOI":"10.3390\/jcdd11070207","article-title":"Machine learning approaches for stroke risk prediction: findings from the Suita study","volume":"11","author":"Vu","year":"2024","journal-title":"J Cardiovasc Dev Dis"},{"key":"ref36","doi-asserted-by":"publisher","first-page":"232","DOI":"10.1038\/s43856-024-00666-w","article-title":"Machine learning for early dynamic prediction of functional outcome after stroke","volume":"4","author":"Klug","year":"2024","journal-title":"Commun Med"},{"key":"ref37","doi-asserted-by":"publisher","first-page":"502","DOI":"10.3389\/fphys.2023.1188502","article-title":"Combining conventional ultrasound and ultrasound elastography to predict HER2 status in patients with breast cancer","volume":"14","author":"Zhuo","year":"2023","journal-title":"Front Physiol"},{"key":"ref38","doi-asserted-by":"publisher","first-page":"2583","DOI":"10.1161\/STROKEAHA.125.052583","article-title":"Revolution in stroke treatment over 50 years and predicting stroke care in 2050","volume":"2025","author":"Broderick","year":"2025","journal-title":"Stroke"},{"key":"ref39","doi-asserted-by":"publisher","first-page":"882","DOI":"10.1177\/23969873251317347","article-title":"First ischemic stroke in young adults: Sex and age-related differences in stroke rates, risk factors, and etiologies","volume":"10","author":"Sturzenegger","year":"2025","journal-title":"Eur Stroke J"},{"key":"ref40","doi-asserted-by":"publisher","first-page":"105","DOI":"10.1161\/STROKEAHA.124.048385","article-title":"Association of Duration of recognized hypertension and stroke risk: the REGARDS study","volume":"56","author":"Howard","year":"2025","journal-title":"Stroke"},{"key":"ref41","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1159\/000549080","article-title":"Association between a body shape index and the prevalence of stroke: a cross-sectional study","volume":"2025","author":"Shi","year":"2025","journal-title":"Neuroepidemiology"},{"key":"ref42","doi-asserted-by":"publisher","first-page":"2369","DOI":"10.1161\/STROKEAHA.123.042525","article-title":"Life\u2019s essential 8 and risk of stroke: a prospective community-based study","volume":"54","author":"Wu","year":"2023","journal-title":"Stroke"},{"key":"ref43","doi-asserted-by":"publisher","first-page":"e034057","DOI":"10.1161\/JAHA.123.034057","article-title":"Incidence and predictors of stroke in Australian adults with congenital heart disease (2000\u20132017)","volume":"13","author":"Sodhi-Berry","year":"2024","journal-title":"J Am Heart Assoc"},{"key":"ref44","doi-asserted-by":"publisher","first-page":"105050","DOI":"10.1016\/j.ijmedinf.2023.105050","article-title":"An interpretable machine learning approach for predicting 30-day readmission after stroke","volume":"174","author":"Lv","year":"2023","journal-title":"Int J Med Inform"}],"container-title":["Frontiers in Neurology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fneur.2025.1716984\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,12]],"date-time":"2026-01-12T05:45:57Z","timestamp":1768196757000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fneur.2025.1716984\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,1,12]]},"references-count":44,"alternative-id":["10.3389\/fneur.2025.1716984"],"URL":"https:\/\/doi.org\/10.3389\/fneur.2025.1716984","relation":{},"ISSN":["1664-2295"],"issn-type":[{"value":"1664-2295","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,1,12]]},"article-number":"1716984"}}