{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,29]],"date-time":"2026-05-29T11:37:12Z","timestamp":1780054632425,"version":"3.54.0"},"reference-count":45,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2024,2,14]],"date-time":"2024-02-14T00:00:00Z","timestamp":1707868800000},"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. Neurosci."],"abstract":"<jats:sec><jats:title>Objective<\/jats:title><jats:p>To establish a deep learning model for the detection of hypoxic\u2013ischemic encephalopathy (HIE) features on CT scans and to compare various networks to determine the best input data format.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>168 head CT scans of patients after cardiac arrest were retrospectively identified and classified into two categories: 88 (52.4%) with radiological evidence of severe HIE and 80 (47.6%) without signs of HIE. These images were randomly divided into a training and a test set, and five deep learning models based on based on Densely Connected Convolutional Networks (DenseNet121) were trained and validated using different image input formats (2D and 3D images).<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>All optimized stacked 2D and 3D networks could detect signs of HIE. The networks based on the data as 2D image data stacks provided the best results (<jats:italic>S100:<\/jats:italic> AUC: 94%, ACC: 79%, <jats:italic>S50:<\/jats:italic> AUC: 93%, ACC: 79%). We provide visual explainability data for the decision making of our AI model using Gradient-weighted Class Activation Mapping.<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusion<\/jats:title><jats:p>Our proof-of-concept deep learning model can accurately identify signs of HIE on CT images. Comparing different 2D- and 3D-based approaches, most promising results were achieved by 2D image stack models. After further clinical validation, a deep learning model of HIE detection based on CT images could be implemented in clinical routine and thus aid clinicians in characterizing imaging data and predicting outcome.<\/jats:p><\/jats:sec>","DOI":"10.3389\/fnins.2024.1245791","type":"journal-article","created":{"date-parts":[[2024,2,14]],"date-time":"2024-02-14T04:16:01Z","timestamp":1707884161000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":7,"title":["Deep learning-enabled detection of hypoxic\u2013ischemic encephalopathy after cardiac arrest in CT scans: a comparative study of 2D and 3D approaches"],"prefix":"10.3389","volume":"18","author":[{"given":"Noah S.","family":"Molinski","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Martin","family":"Kenda","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Christoph","family":"Leithner","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jens","family":"Nee","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Christian","family":"Storm","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael","family":"Scheel","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aymen","family":"Meddeb","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2024,2,14]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"449","DOI":"10.1007\/s10278-017-9983-4","article-title":"Deep learning for brain MRI segmentation: state of the art and future directions","volume":"30","author":"Akkus","year":"2017","journal-title":"J. Digit. Imaging"},{"key":"ref2","doi-asserted-by":"crossref","DOI":"10.1201\/9780429258589","volume-title":"Practical statistics for medical research","author":"Altman","year":"1990"},{"key":"ref3","doi-asserted-by":"publisher","first-page":"519","DOI":"10.1002\/jmri.26643","article-title":"Differentiation between glioblastoma, brain metastasis and subtypes using Radiomics analysis","volume":"50","author":"Artzi","year":"2019","journal-title":"J. Magn. Reson. Imaging"},{"key":"ref4","doi-asserted-by":"publisher","first-page":"5611","DOI":"10.3390\/s20195611","article-title":"Accurate and efficient intracranial hemorrhage detection and subtype classification in 3D CT scans with convolutional and long short-term memory neural networks","volume":"20","author":"Burduja","year":"2020","journal-title":"Sensors"},{"key":"ref5","doi-asserted-by":"publisher","first-page":"414","DOI":"10.1007\/s12028-019-00900-7","article-title":"Interobserver variability in the recognition of hypoxic\u2013ischemic brain injury on computed tomography soon after out-of-hospital cardiac arrest","volume":"33","author":"Caraganis","year":"2020","journal-title":"Neurocrit. Care."},{"key":"ref6","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/ijcnn55064.2022.9892850","article-title":"Are 3D better than 2D convolutional neural networks for medical imaging semantic segmentation?","author":"Crespi","year":"2022","journal-title":"2022 Int. Jt. Conf. Neural Netw."},{"key":"ref7","doi-asserted-by":"publisher","first-page":"1348","DOI":"10.1016\/j.resuscitation.2014.05.022","article-title":"The association between a quantitative computed tomography (CT) measurement of cerebral edema and outcomes in post-cardiac arrest\u2014a validation study","volume":"85","author":"Cristia","year":"2014","journal-title":"Resuscitation"},{"key":"ref8","volume-title":"S1-Leitlinie Hypoxisch-Isch\u00e4mische Enzephalopathie (HIE) Im Erwachsenenalter","year":"2017"},{"key":"ref9","doi-asserted-by":"publisher","first-page":"109762","DOI":"10.1016\/j.resuscitation.2023.109762","article-title":"Paucity of Neuroprognostic testing after cardiac arrest in the United States","volume":"188","author":"Elmer","year":"2023","journal-title":"Resuscitation"},{"key":"ref10","doi-asserted-by":"publisher","first-page":"e2225608","DOI":"10.1001\/jamanetworkopen.2022.25608","article-title":"Development and validation of a deep learning model for brain tumor diagnosis and classification using magnetic resonance imaging","volume":"5","author":"Gao","year":"2022","journal-title":"JAMA Netw. Open"},{"key":"ref11","doi-asserted-by":"publisher","first-page":"e517","DOI":"10.1161\/cir.0000000000000702","article-title":"Standards for studies of neurological prognostication in comatose survivors of cardiac arrest: a scientific statement from the American Heart Association","volume":"140","author":"Geocadin","year":"2019","journal-title":"Circulation"},{"key":"ref12","doi-asserted-by":"publisher","first-page":"14","DOI":"10.3390\/data5010014","article-title":"Intracranial hemorrhage segmentation using a deep convolutional model","volume":"5","author":"Hssayeni","year":"2020","journal-title":"Data"},{"key":"ref13","article-title":"Densely connected convolutional networks","author":"Huang","year":"2016","journal-title":"Arxiv"},{"key":"ref14","doi-asserted-by":"publisher","first-page":"782","DOI":"10.1016\/j.neuroimage.2011.09.015","article-title":"FSL","volume":"62","author":"Jenkinson","year":"2012","journal-title":"NeuroImage"},{"key":"ref15","doi-asserted-by":"publisher","first-page":"118","DOI":"10.1093\/biostatistics\/kxj037","article-title":"Adjusting batch effects in microarray expression data using empirical Bayes methods","volume":"8","author":"Johnson","year":"2007","journal-title":"Biostatistics"},{"key":"ref16","doi-asserted-by":"publisher","first-page":"990208","DOI":"10.3389\/fneur.2022.990208","article-title":"Inter-rater agreement between humans and Computer in Quantitative Assessment of computed tomography after cardiac arrest","volume":"13","author":"Kenda","year":"2022","journal-title":"Front. Neurol."},{"key":"ref17","doi-asserted-by":"publisher","first-page":"e1212","DOI":"10.1097\/ccm.0000000000005198","article-title":"Automated assessment of brain CT after cardiac arrest\u2014an observational derivation\/validation cohort study","volume":"49","author":"Kenda","year":"2021","journal-title":"Crit. Care Med."},{"key":"ref18","doi-asserted-by":"publisher","first-page":"57","DOI":"10.1186\/1757-7241-21-57","article-title":"Early brain computed tomography findings are associated with outcome in patients treated with therapeutic hypothermia after out-of-hospital cardiac arrest","volume":"21","author":"Kim","year":"2013","journal-title":"Scand J. Trauma Resusc. Emerg. Med."},{"key":"ref19","doi-asserted-by":"publisher","first-page":"1227","DOI":"10.2214\/ajr.141.6.1227","article-title":"Early CT findings of global central nervous system Hypoperfusion","volume":"141","author":"Kjos","year":"1983","journal-title":"Am. J. Roentgenol."},{"key":"ref20","doi-asserted-by":"publisher","first-page":"46","DOI":"10.1016\/j.resuscitation.2015.07.027","article-title":"Prognostic values of gray matter to white matter ratios on early brain computed tomography in adult comatose patients after out-of-hospital cardiac arrest of cardiac etiology","volume":"96","author":"Lee","year":"2015","journal-title":"Resuscitation"},{"key":"ref21","doi-asserted-by":"publisher","first-page":"200","DOI":"10.1186\/s12859-020-3503-0","article-title":"A multi-label classification model for full slice brain computerised tomography image","volume":"21","author":"Li","year":"2020","journal-title":"BMC Bioinformat."},{"key":"ref22","doi-asserted-by":"publisher","first-page":"766","DOI":"10.21037\/qims-21-245","article-title":"A phantom study comparing low-dose CT physical image quality from five different CT scanners","volume":"12","author":"Li","year":"2021","journal-title":"Quant. Imaging Med. Surg."},{"key":"ref23","doi-asserted-by":"publisher","first-page":"97","DOI":"10.3389\/fnins.2019.00097","article-title":"A U-net deep learning framework for high performance vessel segmentation in patients with cerebrovascular disease","volume":"13","author":"Livne","year":"2019","journal-title":"Front. Neurosci."},{"key":"ref24","doi-asserted-by":"publisher","first-page":"974","DOI":"10.1007\/s12028-021-01405-y","article-title":"Machine learning for early detection of hypoxic-ischemic brain injury after cardiac arrest","volume":"36","author":"Mansour","year":"2022","journal-title":"Neurocrit. Care."},{"key":"ref25","doi-asserted-by":"publisher","first-page":"950","DOI":"10.3390\/tomography7040078","article-title":"Evaluation of a deep learning algorithm for automated spleen segmentation in patients with conditions directly or indirectly affecting the spleen","volume":"7","author":"Meddeb","year":"2021","journal-title":"Tomogr"},{"key":"ref26","doi-asserted-by":"publisher","first-page":"5476","DOI":"10.3390\/cancers14225476","article-title":"Two-stage deep learning model for automated segmentation and classification of splenomegaly","volume":"14","author":"Meddeb","year":"2022","journal-title":"Cancers"},{"key":"ref27","doi-asserted-by":"publisher","first-page":"1180","DOI":"10.1016\/j.resuscitation.2011.04.001","article-title":"Association between a quantitative CT scan measure of brain edema and outcome after cardiac arrest","volume":"82","author":"Metter","year":"2011","journal-title":"Resuscitation"},{"key":"ref28","doi-asserted-by":"publisher","first-page":"363","DOI":"10.1007\/s12028-022-01526-y","article-title":"Comment on \"machine learning for early detection of hypoxic\u2011ischemic brain injury after cardiac arrest\"","volume":"37","author":"Molinski","year":"2022","journal-title":"Neurocrit. Care."},{"key":"ref29","doi-asserted-by":"publisher","first-page":"e200029","DOI":"10.1148\/ryai.2020200029","article-title":"Checklist for artificial intelligence in medical imaging (CLAIM): a guide for authors and reviewers","volume":"2","author":"Mongan","year":"2020","journal-title":"Radiol. Artif. Intell."},{"key":"ref30","doi-asserted-by":"publisher","first-page":"21","DOI":"10.1016\/j.resuscitation.2018.08.024","article-title":"Gray matter to white matter ratio for predicting neurological outcomes in patients treated with target temperature management after cardiac arrest: a systematic review and meta-analysis","volume":"132","author":"Na","year":"2018","journal-title":"Resuscitation"},{"key":"ref31","doi-asserted-by":"publisher","first-page":"8214","DOI":"10.1038\/s41598-022-11990-3","article-title":"Deep learning model for the automatic classification of COVID-19 pneumonia, non-COVID-19 pneumonia, and the healthy: a multi-center retrospective study","volume":"12","author":"Nishio","year":"2022","journal-title":"Sci. Rep."},{"key":"ref32","doi-asserted-by":"publisher","first-page":"369","DOI":"10.1007\/s00134-021-06368-4","article-title":"European resuscitation council and European Society of Intensive Care Medicine Guidelines 2021: post-resuscitation care","volume":"47","author":"Nolan","year":"2021","journal-title":"Intensiv. Care Med."},{"key":"ref33","doi-asserted-by":"publisher","first-page":"202","DOI":"10.1016\/j.resuscitation.2015.07.018","article-title":"European resuscitation council and European Society of Intensive Care Medicine Guidelines for post-resuscitation care 2015 section 5 of the European resuscitation council guidelines for resuscitation 2015","volume":"95","author":"Nolan","year":"2015","journal-title":"Resuscitation"},{"key":"ref34","doi-asserted-by":"publisher","first-page":"172","DOI":"10.2967\/jnumed.121.262464","article-title":"A guide to ComBat harmonization of imaging biomarkers in multicenter studies","volume":"63","author":"Orlhac","year":"2021","journal-title":"J. Nucl. Med."},{"key":"ref35","first-page":"2825","article-title":"Scikit-learn: machine learning in Python","volume":"12","author":"Pedregosa","year":"2012","journal-title":"J. Mach. Learn. Res."},{"key":"ref36","doi-asserted-by":"publisher","first-page":"350","DOI":"10.1120\/jacmp.v16i2.4972","article-title":"CT image quality over time: comparison of image quality for six different CT scanners over a six-year period","volume":"16","author":"Roa","year":"2015","journal-title":"J. Appl. Clin. M\u00e9d. Phys."},{"key":"ref37","doi-asserted-by":"publisher","first-page":"23","DOI":"10.1186\/1757-7241-21-23","article-title":"The prognostic value of gray-white-matter ratio in cardiac arrest patients treated with hypothermia","volume":"21","author":"Scheel","year":"2013","journal-title":"Scand J. Trauma Resusc. Emerg. Med."},{"key":"ref38","doi-asserted-by":"publisher","first-page":"336","DOI":"10.1007\/s11263-019-01228-7","article-title":"Grad-CAM: visual explanations from deep networks via gradient-based localization","volume":"128","author":"Selvaraju","year":"2016","journal-title":"Arxiv"},{"key":"ref39","doi-asserted-by":"publisher","first-page":"5097","DOI":"10.3390\/s20185097","article-title":"3D deep learning on medical images: a review","volume":"20","author":"Singh","year":"2020","journal-title":"Sensors"},{"key":"ref40","doi-asserted-by":"publisher","first-page":"8","DOI":"10.1016\/j.resuscitation.2019.09.025","article-title":"Timing of brain computed tomography and accuracy of outcome prediction after cardiac arrest","volume":"145","author":"Streitberger","year":"2019","journal-title":"Resuscitation"},{"key":"ref41","doi-asserted-by":"publisher","first-page":"924","DOI":"10.1038\/s41591-022-01772-9","article-title":"Reporting guideline for the early-stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI","volume":"28","author":"Vasey","year":"2022","journal-title":"Nat. Med."},{"key":"ref42","doi-asserted-by":"publisher","first-page":"9","DOI":"10.1186\/s40537-016-0043-6","article-title":"A survey of transfer learning","volume":"3","author":"Weiss","year":"2016","journal-title":"J. Big Data"},{"key":"ref43","doi-asserted-by":"publisher","first-page":"541817","DOI":"10.3389\/fnins.2020.541817","article-title":"Deep network for the automatic segmentation and quantification of intracranial hemorrhage on CT","volume":"14","author":"Xu","year":"2021","journal-title":"Front. Neurosci."},{"key":"ref44","doi-asserted-by":"publisher","first-page":"3997","DOI":"10.1093\/eurheartj\/ehac262","article-title":"ESC guidelines for the Management of Patients with ventricular arrhythmias and the prevention of sudden cardiac death","volume":"43","author":"Zeppenfeld","year":"2022","journal-title":"Eur. Heart J."},{"key":"ref45","doi-asserted-by":"publisher","first-page":"2384830","DOI":"10.1155\/2022\/2384830","article-title":"Dense convolutional network and its application in medical image analysis","volume":"2022","author":"Zhou","year":"2022","journal-title":"Biomed. Res. Int."}],"container-title":["Frontiers in Neuroscience"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnins.2024.1245791\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,2,14]],"date-time":"2024-02-14T04:16:06Z","timestamp":1707884166000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnins.2024.1245791\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,2,14]]},"references-count":45,"alternative-id":["10.3389\/fnins.2024.1245791"],"URL":"https:\/\/doi.org\/10.3389\/fnins.2024.1245791","relation":{},"ISSN":["1662-453X"],"issn-type":[{"value":"1662-453X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,2,14]]},"article-number":"1245791"}}