{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,6]],"date-time":"2026-06-06T05:16:47Z","timestamp":1780723007271,"version":"3.54.1"},"reference-count":28,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2024,8,1]],"date-time":"2024-08-01T00:00:00Z","timestamp":1722470400000},"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. Artif. Intell."],"abstract":"<jats:sec><jats:title>Purpose<\/jats:title><jats:p>Computed Tomography Angiography (CTA) is the first line of imaging in the diagnosis of Large Vessel Occlusion (LVO) strokes. We trained and independently validated end-to-end automated deep learning pipelines to predict 3-month outcomes after anterior circulation LVO thrombectomy based on admission CTAs.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>We split a dataset of 591 patients into training\/cross-validation (n\u2009=\u2009496) and independent test set (n\u2009=\u200995). We trained separate models for outcome prediction based on admission \u201cCTA\u201d images alone, \u201cCTA + Treatment\u201d (including time to thrombectomy and reperfusion success information), and \u201cCTA + Treatment\u2009 + Clinical\u201d (including admission age, sex, and NIH stroke scale). A binary (favorable) outcome was defined based on a 3-month modified Rankin Scale \u2264 2. The model was trained on our dataset based on the pre-trained ResNet-50 3D Convolutional Neural Network (\u201cMedicalNet\u201d) and included CTA preprocessing steps.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>We generated an ensemble model from the 5-fold cross-validation, and tested it in the independent test cohort, with receiver operating characteristic area under the curve (AUC, 95% confidence interval) of 70 (0.59\u20130.81) for \u201cCTA,\u201d 0.79 (0.70\u20130.89) for \u201cCTA + Treatment,\u201d and 0.86 (0.79\u20130.94) for \u201cCTA + Treatment + Clinical\u201d input models. A \u201cTreatment\u2009+\u2009Clinical\u201d logistic regression model achieved an AUC of 0.86 (0.79\u20130.93).<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusion<\/jats:title><jats:p>Our results show the feasibility of an end-to-end automated model to predict outcomes from admission and post-thrombectomy reperfusion success. Such a model can facilitate prognostication in telehealth transfer and when a thorough neurological exam is not feasible due to language barrier or pre-existing morbidities.<\/jats:p><\/jats:sec>","DOI":"10.3389\/frai.2024.1369702","type":"journal-article","created":{"date-parts":[[2024,8,1]],"date-time":"2024-08-01T05:10:57Z","timestamp":1722489057000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":9,"title":["Deep learning for prediction of post-thrombectomy outcomes based on admission CT angiography in large vessel occlusion stroke"],"prefix":"10.3389","volume":"7","author":[{"given":"Jakob","family":"Sommer","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fiona","family":"Dierksen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tal","family":"Zeevi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Anh Tuan","family":"Tran","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Emily W.","family":"Avery","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Adrian","family":"Mak","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ajay","family":"Malhotra","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Charles C.","family":"Matouk","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Guido J.","family":"Falcone","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Victor","family":"Torres-Lopez","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sanjey","family":"Aneja","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"James","family":"Duncan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lauren H.","family":"Sansing","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kevin N.","family":"Sheth","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Seyedmehdi","family":"Payabvash","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2024,8,1]]},"reference":[{"key":"ref1","author":"Alexandari","year":""},{"key":"ref2","doi-asserted-by":"publisher","first-page":"2033","DOI":"10.1016\/j.neuroimage.2010.09.025","article-title":"A reproducible evaluation of Ants similarity metric performance in brain image registration","volume":"54","author":"Avants","year":"2011","journal-title":"Neuro Image"},{"key":"ref3","doi-asserted-by":"publisher","first-page":"103034","DOI":"10.1016\/j.nicl.2022.103034","article-title":"Ct angiographic radiomics signature for risk stratification in anterior large vessel occlusion stroke","volume":"34","author":"Avery","year":"2022","journal-title":"Neuroimage Clin."},{"key":"ref4","doi-asserted-by":"publisher","first-page":"1753","DOI":"10.1016\/S0140-6736(23)02032-9","article-title":"Endovascular thrombectomy for acute ischaemic stroke with established large infarct: multicentre, open-label, randomised trial","volume":"402","author":"Bendszus","year":"2023","journal-title":"Lancet"},{"key":"ref5","doi-asserted-by":"publisher","first-page":"541","DOI":"10.1148\/radiol.2442061028","article-title":"Acute brain infarct: detection and delineation with Ct angiographic source images versus nonenhanced Ct scans","volume":"244","author":"Camargo","year":"2007","journal-title":"Radiology"},{"key":"ref6","volume-title":"Monai: An open-source framework for deep learning in healthcare. [Preprint].","author":"Cardoso","year":"2022"},{"key":"ref7","volume-title":"Med3D: Transfer learning for 3D medical image analysis. Corr, abs\/1904.00625","author":"Chen","year":"2019"},{"key":"ref8","doi-asserted-by":"publisher","first-page":"e37595","DOI":"10.7759\/cureus.37595","article-title":"Reliability of the National Institutes of Health (Nih) stroke scale between emergency room and neurology physicians for initial stroke severity scoring","volume":"15","author":"Cummock","year":"2023","journal-title":"Cureus"},{"key":"ref9","doi-asserted-by":"publisher","first-page":"837","DOI":"10.2307\/2531595","article-title":"Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach","volume":"44","author":"Delong","year":"1988","journal-title":"Biometrics"},{"key":"ref14","article-title":"M3d-CAM: A PyTorch library to generate 3D data attention maps for medical deep learning","author":"Gotkowski","year":"2020","journal-title":"arXiv [Preprint]."},{"key":"ref10","doi-asserted-by":"publisher","first-page":"103516","DOI":"10.1016\/j.compbiomed.2019.103516","article-title":"Data-efficient deep learning of radiological image data for outcome prediction after endovascular treatment of patients with acute ischemic stroke","volume":"115","author":"Hilbert","year":"2019","journal-title":"Comput. Biol. Med."},{"key":"ref11","doi-asserted-by":"publisher","first-page":"1272","DOI":"10.1056\/NEJMoa2213379","article-title":"Trial of endovascular therapy for acute ischemic stroke with large infarct","volume":"388","author":"Huo","year":"2023","journal-title":"N. Engl. J. Med."},{"key":"ref12","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1186\/s40708-020-00104-2","article-title":"Gan-based synthetic brain pet image generation","volume":"7","author":"Islam","year":"2020","journal-title":"Brain Inform."},{"key":"ref13","doi-asserted-by":"publisher","first-page":"110543","DOI":"10.1016\/j.knosys.2023.110543","article-title":"Improving generalization of convolutional neural network through contrastive augmentation","volume":"272","author":"Li","year":"2023","journal-title":"Knowl.-Based Syst."},{"key":"ref15","doi-asserted-by":"publisher","first-page":"3604","DOI":"10.3390\/diagnostics13243604","article-title":"Deep learning-based extraction of biomarkers for the prediction of the functional outcome of ischemic stroke patients","volume":"13","author":"Oliveira","year":"2023","journal-title":"Diagnostics"},{"key":"ref16","doi-asserted-by":"publisher","first-page":"957","DOI":"10.1016\/j.neuroimage.2012.03.020","article-title":"Age-specific Ct and Mri templates for spatial normalization","volume":"61","author":"Rorden","year":"2012","journal-title":"NeuroImage"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"940","DOI":"10.1136\/neurintsurg-2016-012628","article-title":"Ct angiography-based collateral flow and time to reperfusion are strong predictors of outcome in endovascular treatment of patients with stroke","volume":"9","author":"Sallustio","year":"2017","journal-title":"J. Neurointerv. Surg."},{"key":"ref18","doi-asserted-by":"publisher","first-page":"1403","DOI":"10.2214\/AJR.19.22347","article-title":"Data augmentation and transfer learning to improve generalizability of an automated prostate segmentation model","volume":"215","author":"Sanford","year":"2020","journal-title":"Am. J. Roentgenol."},{"key":"ref19","doi-asserted-by":"publisher","first-page":"1259","DOI":"10.1056\/NEJMoa2214403","article-title":"Trial of endovascular Thrombectomy for large ischemic strokes","volume":"388","author":"Sarraj","year":"2023","journal-title":"N. Engl. J. Med."},{"key":"ref20","doi-asserted-by":"publisher","first-page":"1652","DOI":"10.1161\/01.STR.0000131271.54098.22","article-title":"Comparison of perfusion computed tomography and computed tomography angiography source images with perfusion-weighted imaging and diffusion-weighted imaging in patients with acute stroke of less than 6 hours\u2019 duration","volume":"35","author":"Schramm","year":"2004","journal-title":"Stroke"},{"key":"ref21","doi-asserted-by":"publisher","first-page":"403","DOI":"10.1007\/s12021-020-09493-5","article-title":"3D deep neural network segmentation of intracerebral hemorrhage: development and validation for clinical trials","volume":"19","author":"Sharrock","year":"2021","journal-title":"Neuroinformatics"},{"key":"ref22","doi-asserted-by":"publisher","first-page":"014","DOI":"10.1385\/NCC:4:1:014","article-title":"Prognostic significance of angiographically confirmed large vessel intracranial occlusion in patients presenting with acute brain ischemia","volume":"4","author":"Smith","year":"2006","journal-title":"Neurocrit. Care."},{"key":"ref23","doi-asserted-by":"publisher","first-page":"258","DOI":"10.1159\/000441088","article-title":"The prognostic value of Ct angiography and Ct perfusion in acute ischemic stroke","volume":"40","author":"Van Seeters","year":"2015","journal-title":"Cerebrovasc. Dis."},{"key":"ref24","doi-asserted-by":"publisher","first-page":"104504","DOI":"10.1016\/j.jstrokecerebrovasdis.2019.104504","article-title":"Large vessel occlusion in acute ischemic stroke patients: a dual-center estimate based on a broad definition of occlusion site","volume":"29","author":"Waqas","year":"2020","journal-title":"J. Stroke Cerebrovasc. Dis."},{"key":"ref25","doi-asserted-by":"publisher","first-page":"1303","DOI":"10.1056\/NEJMoa2118191","article-title":"Endovascular therapy for acute stroke with a large ischemic region","volume":"386","author":"Yoshimura","year":"2022","journal-title":"N. Engl. J. Med."},{"key":"ref26","doi-asserted-by":"publisher","first-page":"2650","DOI":"10.1161\/STROKEAHA.113.001972","article-title":"Recommendations on angiographic revascularization grading standards for acute ischemic stroke: a consensus statement","volume":"44","author":"Zaidat","year":"2013","journal-title":"Stroke"},{"key":"ref27","doi-asserted-by":"publisher","first-page":"110959","DOI":"10.1016\/j.ejrad.2023.110959","article-title":"Non-contrast Ct radiomics and machine learning for outcomes prediction of patients with acute ischemic stroke receiving conventional treatment","volume":"165","author":"Zhang","year":"2023","journal-title":"Eur. J. Radiol."},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1016\/j.metrad.2023.100003","article-title":"A review of uncertainty estimation and its application in medical imaging","volume":"100003","author":"Zou","year":"2023","journal-title":"Meta Radiol."}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2024.1369702\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,8,1]],"date-time":"2024-08-01T05:11:02Z","timestamp":1722489062000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2024.1369702\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,8,1]]},"references-count":28,"alternative-id":["10.3389\/frai.2024.1369702"],"URL":"https:\/\/doi.org\/10.3389\/frai.2024.1369702","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,8,1]]},"article-number":"1369702"}}