{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,16]],"date-time":"2026-11-16T19:12:44Z","timestamp":1794856364234,"version":"4.9.1"},"reference-count":15,"publisher":"Ovid Technologies (Wolters Kluwer Health)","issue":"5","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Invest Radiol"],"published-print":{"date-parts":[[2017,5]]},"abstract":"<jats:sec>\n            <jats:title>Objectives<\/jats:title>\n            <jats:p>Convolutional neural networks (CNNs) are a subtype of artificial neural network that have shown strong performance in computer vision tasks including image classification. To date, there has been limited application of CNNs to chest radiographs, the most frequently performed medical imaging study. We hypothesize CNNs can learn to classify frontal chest radiographs according to common findings from a sufficiently large data set.<\/jats:p>\n          <\/jats:sec>\n          <jats:sec>\n            <jats:title>Materials and Methods<\/jats:title>\n            <jats:p>Our institution's research ethics board approved a single-center retrospective review of 35,038 adult posterior-anterior chest radiographs and final reports performed between 2005 and 2015 (56% men, average age of 56, patient type: 24% inpatient, 39% outpatient, 37% emergency department) with a waiver for informed consent. The GoogLeNet CNN was trained using 3 graphics processing units to automatically classify radiographs as normal (n = 11,702) or into 1 or more of cardiomegaly (n = 9240), consolidation (n = 6788), pleural effusion (n = 7786), pulmonary edema (n = 1286), or pneumothorax (n = 1299). The network's performance was evaluated using receiver operating curve analysis on a test set of 2443 radiographs with the criterion standard being board-certified radiologist interpretation.<\/jats:p>\n          <\/jats:sec>\n          <jats:sec>\n            <jats:title>Results<\/jats:title>\n            <jats:p>Using 256 \u00d7 256-pixel images as input, the network achieved an overall sensitivity and specificity of 91% with an area under the curve of 0.964 for classifying a study as normal (n = 1203). For the abnormal categories, the sensitivity, specificity, and area under the curve, respectively, were 91%, 91%, and 0.962 for pleural effusion (n = 782), 82%, 82%, and 0.868 for pulmonary edema (n = 356), 74%, 75%, and 0.850 for consolidation (n = 214), 81%, 80%, and 0.875 for cardiomegaly (n = 482), and 78%, 78%, and 0.861 for pneumothorax (n = 167).<\/jats:p>\n          <\/jats:sec>\n          <jats:sec>\n            <jats:title>Conclusions<\/jats:title>\n            <jats:p>Current deep CNN architectures can be trained with modest-sized medical data sets to achieve clinically useful performance at detecting and excluding common pathology on chest radiographs.<\/jats:p>\n          <\/jats:sec>","DOI":"10.1097\/rli.0000000000000341","type":"journal-article","created":{"date-parts":[[2016,12,6]],"date-time":"2016-12-06T14:22:53Z","timestamp":1481034173000},"page":"281-287","source":"Crossref","is-referenced-by-count":230,"title":["Training and Validating a Deep Convolutional Neural Network for Computer-Aided Detection and Classification of Abnormalities on Frontal Chest Radiographs"],"prefix":"10.1097","volume":"52","author":[{"given":"Mark","family":"Cicero","sequence":"first","affiliation":[{"name":"From the *Department of Medical Imaging, St Michael's Hospital, and \u2020Department of Pharmaceutical Sciences, University of Toronto, Toronto, Ontario, Canada."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alexander","family":"Bilbily","sequence":"additional","affiliation":[{"name":"From the *Department of Medical Imaging, St Michael's Hospital, and \u2020Department of Pharmaceutical Sciences, University of Toronto, Toronto, Ontario, Canada."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Errol","family":"Colak","sequence":"additional","affiliation":[{"name":"From the *Department of Medical Imaging, St Michael's Hospital, and \u2020Department of Pharmaceutical Sciences, University of Toronto, Toronto, Ontario, Canada."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tim","family":"Dowdell","sequence":"additional","affiliation":[{"name":"From the *Department of Medical Imaging, St Michael's Hospital, and \u2020Department of Pharmaceutical Sciences, University of Toronto, Toronto, Ontario, Canada."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bruce","family":"Gray","sequence":"additional","affiliation":[{"name":"From the *Department of Medical Imaging, St Michael's Hospital, and \u2020Department of Pharmaceutical Sciences, University of Toronto, Toronto, Ontario, Canada."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kuhan","family":"Perampaladas","sequence":"additional","affiliation":[{"name":"From the *Department of Medical Imaging, St Michael's Hospital, and \u2020Department of Pharmaceutical Sciences, University of Toronto, Toronto, Ontario, Canada."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Joseph","family":"Barfett","sequence":"additional","affiliation":[{"name":"From the *Department of Medical Imaging, St Michael's Hospital, and \u2020Department of Pharmaceutical Sciences, University of Toronto, Toronto, Ontario, Canada."}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"276","reference":[{"key":"bib1-20231021","first-page":"1106","article-title":"ImageNet classification with deep convolutional neural networks","volume":"25","year":"2012","journal-title":"Adv Neural Inf Process Syst"},{"key":"bib2-20231021","article-title":"Deep residual learning for image recognition","year":"2015","journal-title":"arXiv preprint arXiv"},{"key":"bib3-20231021","first-page":"1","article-title":"Going deeper with convolutions","year":"2015","journal-title":"In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition"},{"key":"bib4-20231021","first-page":"1","article-title":"Learning representations by back-propagating errors","volume":"5","year":"1988","journal-title":"Cogn Model"},{"key":"bib5-20231021","first-page":"1","article-title":"The \u2018big data\u2019 revolution in healthcare\u2014accelerating value and innovation","year":"2013","journal-title":"Center for US Health System Reform Business Technology Office"},{"key":"bib7-20231021","doi-asserted-by":"crossref","first-page":"261","DOI":"10.1097\/RLI.0000000000000127","article-title":"Automated classification of usual interstitial pneumonia using regional volumetric texture analysis in high-resolution computed tomography","volume":"50","year":"2015","journal-title":"Invest Radiol"},{"key":"bib8-20231021","first-page":"9","article-title":"Practical issues in calculating the sample size for prevalence studies","volume":"1","year":"2006","journal-title":"Med Stat Arch Orofacial Sci"},{"key":"bib9-20231021","article-title":"Intriguing properties of neural networks","year":"2014","journal-title":"arXiv"},{"key":"bib10-20231021","article-title":"Rich feature hierarchies for accurate object detection and semantic segmentation","year":"2013","journal-title":"arXiv"},{"key":"bib11-20231021","first-page":"5375","article-title":"Learning deep representation for imbalanced classification","year":"2016","journal-title":"In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition"},{"key":"bib12-20231021","doi-asserted-by":"crossref","first-page":"102","DOI":"10.1136\/bmj.309.6947.102","article-title":"Diagnostic tests 2: predictive values","volume":"309","year":"1994","journal-title":"BMJ"},{"key":"bib13-20231021","article-title":"Deep adaptive network: an efficient deep neural network with sparse binary connections","year":"2016","journal-title":"arXiv preprint arXiv"},{"key":"bib14-20231021","article-title":"A new method to visualize deep neural networks","year":"2016","journal-title":"arXiv"},{"key":"bib15-20231021","article-title":"Faster R-CNN: towards real-time object detection with region proposal networks","year":"2016","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"bib16-20231021","doi-asserted-by":"crossref","first-page":"1182","DOI":"10.1109\/TMI.2016.2528129","article-title":"Automatic detection of cerebral microbleeds from MR images via 3D convolutional neural networks","volume":"35","year":"2016","journal-title":"IEEE Trans Med Imaging"}],"container-title":["Investigative Radiology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/journals.lww.com\/00004424-201705000-00004","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,10,21]],"date-time":"2023-10-21T05:18:14Z","timestamp":1697865494000},"score":1,"resource":{"primary":{"URL":"https:\/\/journals.lww.com\/00004424-201705000-00004"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2017,5]]},"references-count":15,"journal-issue":{"issue":"5","published-print":{"date-parts":[[2017]]}},"URL":"https:\/\/doi.org\/10.1097\/rli.0000000000000341","relation":{},"ISSN":["1536-0210","0020-9996"],"issn-type":[{"value":"1536-0210","type":"electronic"},{"value":"0020-9996","type":"print"}],"subject":[],"published":{"date-parts":[[2017,5]]}}}