{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,24]],"date-time":"2026-07-24T05:02:18Z","timestamp":1784869338913,"version":"3.55.0"},"reference-count":36,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,7,24]],"date-time":"2026-07-24T00:00:00Z","timestamp":1784851200000},"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>Ischemic stroke is a leading global cause of death and disability. Accurate, rapid infarct territory localization is critical for timely treatment, but ultra-early ischemic changes on non-contrast CT (NCCT) are subtle and highly physician-dependent.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>This retrospective study enrolled acute ischemic stroke patients with single-territory infarction. Clinical data and NCCT images were collected. Quantitative imaging features (HU values, infarct volume, spatial distribution) were extracted from vascular perfusion territories. A strict leak-proof pipeline was adopted: dataset stratification, missing value imputation, scaling, and Synthetic Minority Oversampling Technique (SMOTE) were applied only to the training set, full feature retention (VIF\u202f&amp;lt;\u202f5), and Logistic Regression (LR) and XGBoost models were constructed. Standardized labeling was performed. Anterior cerebral artery (ACA), middle cerebral artery (MCA), and posterior cerebral artery (PCA) territories were defined by two neuroradiologists using diffusion-weighted imaging (DWI) (gold standard, Kappa analysis). Multi-territory infarcts were excluded. TotalSegmentator segmented parenchymal perfusion territories. DWI infarct ROIs were registered to NCCT. Model performance was assessed by Area under the receiver operating characteristic curve (AUC), sensitivity, specificity, precision and calibration (Brier score, slope), with DeLong-Bonferroni tests. Robust validation used 5-fold repeated stratified cross-validation (10 repetitions).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      A total of 118 eligible patients were enrolled (median age: 72.0\u202fyears; IQR: 65.0\u201379.0; 85.6% male). All patients received NCCT within 24\u202fh of stroke onset, with an average NIHSS score of 6.05\u202f\u00b1\u202f3.26. Inter-rater Kappa for infarct labeling reached 0.89 (excellent agreement). On the test set, XGBoost yielded AUCs of 0.92 (ACA), 0.87 (MCA), and 0.95 (PCA), with a cross-validated mean AUC of 0.94\u202f\u00b1\u202f0.03. Mild overfitting was noted due to a performance drop from training to test data. LR produced AUCs ranging 0.78\u20130.86 with balanced sensitivity. For left\/right\/bilateral infarct classification, XGBoost achieved a higher mean AUC (0.90 vs. 0.82,\n                      <jats:italic>p<\/jats:italic>\n                      \u202f&amp;lt;\u202f0.01). Both models exhibited good calibration.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>XGBoost demonstrates excellent diagnostic performance for infarct territory localization, whereas LR offers stable sensitivity. Strict protocols and robust validation support its clinical auxiliary value. Future studies will incorporate multi-center external validation and multi-territory infarcts to improve generalizability.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fneur.2026.1794563","type":"journal-article","created":{"date-parts":[[2026,7,24]],"date-time":"2026-07-24T04:23:36Z","timestamp":1784867016000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Machine learning-based prediction of cerebral artery territorial infarcts in acute ischemic stroke using non-contrast CT"],"prefix":"10.3389","volume":"17","author":[{"given":"Yang","family":"Guo","sequence":"first","affiliation":[{"name":"Department of Radiology, Peking University Aerospace Clinical Medical College","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shui","family":"Liu","sequence":"additional","affiliation":[{"name":"Department of Radiology, Aerospace Center Hospital","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hongmei","family":"Sun","sequence":"additional","affiliation":[{"name":"Department of Neurology, Aerospace Center Hospital","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiukun","family":"Jin","sequence":"additional","affiliation":[{"name":"Department of Neurology, Aerospace Center Hospital","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yan","family":"Zhang","sequence":"additional","affiliation":[{"name":"Department of Neurology, Aerospace Center Hospital","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jian","family":"Sun","sequence":"additional","affiliation":[{"name":"Department of Radiology, Aerospace Center Hospital","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhiqun","family":"Wang","sequence":"additional","affiliation":[{"name":"Department of Neurology, Aerospace Center Hospital","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lvming","family":"Zhang","sequence":"additional","affiliation":[{"name":"Department of Neurology, Aerospace Center Hospital","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,7,24]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"1250","DOI":"10.1001\/jamaneurol.2024.3659","article-title":"Mobile stroke unit management in patients with acute ischemic stroke eligible for intravenous thrombolysis","volume":"81","author":"Mac Grory","year":"2024","journal-title":"JAMA Neurol"},{"key":"ref2","doi-asserted-by":"publisher","first-page":"100345","DOI":"10.1016\/j.xgen.2023.100345","article-title":"Multi-ancestry meta-analysis identifies 5 novel loci for ischemic stroke and reveals heterogeneity of effects between sexes and ancestries","volume":"3","author":"Surakka","year":"2023","journal-title":"Cell Genom"},{"key":"ref3","doi-asserted-by":"publisher","first-page":"e137","DOI":"10.1212\/WNL.0000000000207387","article-title":"Global burden, risk factor analysis, and prediction study of ischemic stroke, 1990\u20132030","volume":"101","author":"Fan","year":"1990","journal-title":"Neurology"},{"key":"ref4","doi-asserted-by":"publisher","first-page":"394","DOI":"10.1001\/jamaneurol.2024.0190","article-title":"Burden of ischemic and hemorrhagic stroke across the US from 1990 to 2019","volume":"81","author":"Renedo","year":"1990","journal-title":"JAMA Neurol"},{"key":"ref5","doi-asserted-by":"publisher","first-page":"e040493","DOI":"10.1161\/JAHA.124.040493","article-title":"Survivors of stroke face greater social risks than adults who have not had a stroke: results from a nationwide study","volume":"14","author":"Skolarus","year":"2025","journal-title":"J Am Heart Assoc"},{"key":"ref6","doi-asserted-by":"publisher","first-page":"e0322606","DOI":"10.1371\/journal.pone.0322606","article-title":"Global burden of ischemic stroke in adults aged 60 years and older from 1990 to 2021: a population-based study","volume":"20","author":"Chen","year":"2025","journal-title":"PLoS One"},{"key":"ref7","doi-asserted-by":"publisher","first-page":"e0322588","DOI":"10.1371\/journal.pone.0322588","article-title":"Impact of surgery on rehabilitation care and quality of life perceived by patients with post-stroke upper limb spasticity: study protocol for a randomised controlled trial","volume":"20","author":"Hurtado-Olmo","year":"2025","journal-title":"PLoS One"},{"key":"ref8","doi-asserted-by":"publisher","first-page":"954","DOI":"10.1161\/CIRCRESAHA.123.323371","article-title":"LOX-1 and MMP-9 inhibition attenuates the detrimental effects of delayed rt-PA therapy and improves outcomes after acute ischemic stroke","volume":"134","author":"Arkelius","year":"2024","journal-title":"Circ Res"},{"key":"ref9","doi-asserted-by":"publisher","first-page":"2348","DOI":"10.1093\/brain\/awaf013","article-title":"Prediction of tissue and clinical thrombectomy outcome in acute ischaemic stroke using deep learning","volume":"148","author":"von Braun","year":"2025","journal-title":"Brain"},{"key":"ref10","doi-asserted-by":"publisher","first-page":"1017","DOI":"10.1002\/ana.26925","article-title":"Current state of evidence for neuroimaging paradigms in management of acute ischemic stroke","volume":"95","author":"Sarraj","year":"2024","journal-title":"Ann Neurol"},{"key":"ref11","doi-asserted-by":"publisher","first-page":"504","DOI":"10.22541\/ccm.2024.08.0504","article-title":"Outcomes and complications associated with mechanical thrombectomy in the treatment of acute ischemic stroke","volume":"8","author":"Ahmed","year":"2024","journal-title":"Cardiol Cardiovasc Med"},{"key":"ref12","doi-asserted-by":"publisher","first-page":"e40475","DOI":"10.1097\/MD.0000000000040475","article-title":"A nomogram for predicting hemorrhagic transformation in acute ischemic stroke receiving intravenous thrombolysis with extended time window","volume":"103","author":"Wei","year":"2024","journal-title":"Medicine (Baltimore)"},{"key":"ref13","doi-asserted-by":"publisher","first-page":"2303","DOI":"10.1109\/tmi.2024.3362879","article-title":"Hybrid CNN-transformer network with circular feature interaction for acute ischemic stroke lesion segmentation on non-contrast CT scans","volume":"43","author":"Kuang","year":"2024","journal-title":"IEEE Trans Med Imaging"},{"key":"ref14","doi-asserted-by":"publisher","first-page":"1405","DOI":"10.1007\/s13760-023-02239-5","article-title":"Proposal of multimodal CT-based scoring system in prediction of hemorrhagic transformation in acute ischemic stroke","volume":"123","author":"Elsaid","year":"2023","journal-title":"Acta Neurol Belg"},{"key":"ref15","doi-asserted-by":"publisher","first-page":"2192","DOI":"10.15562\/bmj.v10i1.2192","article-title":"The correlation between C-reactive protein, vascular cell adhesion Molecule-1, and S100b with alberta stroke program early CT score in non-hemorrhagic stroke patients","volume":"10","author":"Retnoningrum","year":"2021","journal-title":"Bali Med J"},{"key":"ref16","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1097\/01.CDR.0000512056.38768.8a","article-title":"Neurovascular emergencies: imaging diagnosis and neurointerventional therapy","volume":"40","author":"Rahman","year":"2017","journal-title":"Contemp Neurosurg"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"189","DOI":"10.4103\/meajo.meajo_154_24","article-title":"Applications of artificial intelligence in choroid visualization for myopia: a comprehensive scoping review","volume":"30","author":"Alhalafi","year":"2023","journal-title":"Middle East Afr J Ophthalmol"},{"key":"ref18","doi-asserted-by":"publisher","first-page":"IA04","DOI":"10.1158\/1557-3265.ADVPRECMED20-IA04","article-title":"Clinical challenges in oncologic imaging: AI support from image analysis to integrated diagnostics","volume":"26","author":"Sala","year":"2020","journal-title":"Clin Cancer Res"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"1171","DOI":"10.1007\/s10554-020-02099-7","article-title":"Multiclass machine learning vs. conventional calculators for stroke\/CVD risk assessment using carotid plaque predictors with coronary angiography scores as gold standard: a 500-participant study","volume":"37","author":"Jamthikar","year":"2021","journal-title":"Int J Cardiovasc Imaging"},{"key":"ref20","doi-asserted-by":"publisher","first-page":"967","DOI":"10.3390\/bioengineering10080967","article-title":"Non-contrasted CT quantitative imaging features for SAH prognosis prediction","volume":"10","author":"Shan","year":"2023","journal-title":"Bioengineering (Basel)"},{"key":"ref21","doi-asserted-by":"publisher","first-page":"889","DOI":"10.3389\/fneur.2020.00889","article-title":"Functional outcome prediction in ischemic stroke: a comparison of machine learning algorithms and regression models","volume":"11","author":"Alaka","year":"2020","journal-title":"Front Neurol"},{"key":"ref22","doi-asserted-by":"publisher","first-page":"284","DOI":"10.1186\/s12871-022-01827-x","article-title":"Machine learning prediction of postoperative major adverse cardiovascular events in geriatric patients: a prospective cohort study","volume":"22","author":"Peng","year":"2022","journal-title":"BMC Anesthesiol"},{"key":"ref23","doi-asserted-by":"publisher","first-page":"W1","DOI":"10.7326\/M18-1377","article-title":"PROBAST: a tool to assess risk of bias and applicability of prediction model studies: explanation and elaboration","volume":"170","author":"Moons","year":"2019","journal-title":"Ann Intern Med"},{"key":"ref24","doi-asserted-by":"publisher","first-page":"e3307","DOI":"10.1002\/brb3.3307","article-title":"Ischemic stroke prediction using machine learning in elderly Chinese population: the Rugao longitudinal ageing study","volume":"13","author":"Chang","year":"2023","journal-title":"Brain Behav"},{"key":"ref25","doi-asserted-by":"publisher","first-page":"e068045","DOI":"10.1136\/bmjopen-2022-068045","article-title":"Development of rapid and effective risk prediction models for stroke in the Chinese population: a cross-sectional study","volume":"13","author":"Qiu","year":"2023","journal-title":"BMJ Open"},{"key":"ref26","doi-asserted-by":"publisher","first-page":"1368902","DOI":"10.3389\/fneur.2024.1368902","article-title":"Prediction of early neurologic deterioration in patients with perforating artery territory infarction using machine learning: a retrospective study","volume":"15","author":"Liu","year":"2024","journal-title":"Front Neurol"},{"key":"ref27","doi-asserted-by":"publisher","first-page":"291","DOI":"10.1007\/s12028-022-01513-3","article-title":"Natural language processing of radiology reports to detect complications of ischemic stroke","volume":"37","author":"Miller","year":"2022","journal-title":"Neurocrit Care"},{"key":"ref28","doi-asserted-by":"publisher","first-page":"e230024","DOI":"10.1148\/ryai.230024","article-title":"TotalSegmentator: robust segmentation of 104 anatomic structures in CT images","volume":"5","author":"Wasserthal","year":"2023","journal-title":"Radiol Artif Intell"},{"key":"ref29","doi-asserted-by":"publisher","first-page":"e230235","DOI":"10.1148\/ryai.230235","article-title":"Totalsegmentator: a gift to the biomedical imaging community","volume":"5","author":"Sebro","year":"2023","journal-title":"Radiol Artif Intell"},{"key":"ref30","doi-asserted-by":"publisher","first-page":"170","DOI":"10.3969\/j.issn.0253-9802.2024.03.004","article-title":"Prognostic prediction model of acute perforating artery occlusion cerebral infarction based on machine learning","volume":"55","author":"Liu","year":"2024","journal-title":"Xin Yi Xue (New Med)"},{"key":"ref31","doi-asserted-by":"publisher","first-page":"e0213653","DOI":"10.1371\/journal.pone.0213653","article-title":"Cardiovascular disease risk prediction using automated machine learning: a prospective study of 423,604 UK biobank participants","volume":"14","author":"Alaa","year":"2019","journal-title":"PLoS One"},{"key":"ref32","doi-asserted-by":"publisher","first-page":"972","DOI":"10.1016\/j.jacasi.2024.09.010","article-title":"Validating machine learning models against the saline test gold standard for primary aldosteronism diagnosis","volume":"4","author":"Liu","year":"2024","journal-title":"JACC Asia"},{"key":"ref33","doi-asserted-by":"publisher","first-page":"e016357","DOI":"10.1161\/JAHA.120.016357","article-title":"Machine learning for stroke risk prediction in Taiwanese adults","volume":"9","author":"Chang","year":"2020","journal-title":"J Am Heart Assoc"},{"key":"ref34","doi-asserted-by":"publisher","first-page":"12667","DOI":"10.1038\/s41598-021-92149-5","article-title":"Development and validation of a novel stroke risk prediction model using machine learning","volume":"11","author":"Chun","year":"2021","journal-title":"Sci Rep"},{"key":"ref35","doi-asserted-by":"publisher","first-page":"2620","DOI":"10.1161\/STROKEAHA.114.005023","article-title":"Risk score to predict hospital-acquired pneumonia after spontaneous intracerebral hemorrhage","volume":"44","author":"Ji","year":"2013","journal-title":"Stroke"},{"key":"ref36","doi-asserted-by":"publisher","first-page":"1263","DOI":"10.1161\/STROKEAHA.118.024293","article-title":"Machine learning-based model for prediction of outcomes in acute stroke","volume":"50","author":"Heo","year":"2019","journal-title":"Stroke"}],"container-title":["Frontiers in Neurology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fneur.2026.1794563\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,24]],"date-time":"2026-07-24T04:23:38Z","timestamp":1784867018000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fneur.2026.1794563\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,24]]},"references-count":36,"alternative-id":["10.3389\/fneur.2026.1794563"],"URL":"https:\/\/doi.org\/10.3389\/fneur.2026.1794563","relation":{},"ISSN":["1664-2295"],"issn-type":[{"value":"1664-2295","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,24]]},"article-number":"1794563"}}