{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,19]],"date-time":"2026-08-19T14:35:22Z","timestamp":1787150122751,"version":"build-2736575974"},"reference-count":36,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,8,19]],"date-time":"2026-08-19T00:00:00Z","timestamp":1787097600000},"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. Radiol."],"abstract":"\n Objectives<\/jats:title>\n To develop a contrastive learning model for lung disease classification using discriminative CT imaging embeddings.<\/jats:p>\n <\/jats:sec>\n \n Methods<\/jats:title>\n \n A total of 1,187 subjects were included: asthma (\n n<\/jats:italic>\n \u2009=\u2009315), COPD (\n n<\/jats:italic>\n \u2009=\u2009355), post-COVID-19 (\n n<\/jats:italic>\n \u2009=\u2009375), and healthy controls (\n n<\/jats:italic>\n \u2009=\u2009142). Of these, 1,003 subjects had a single visit with similarly protocoled CT scans acquired at total lung capacity (TLC) and residual volume (RV), and 92 (33 asthma and 59 post-COVID-19) completed a follow-up visit, with two scans per visit. We developed a modified contrastive learning model incorporating an expert-conditioned routing network and adaptive temperature scaling to learn discriminative embeddings. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC). The embeddings were further validated via k-means clustering, and quantitative CT (qCT) metrics were compared across the derived clusters. The embedding space was used to track disease progression or improvement in the follow-up disease subgroup and to evaluate the model's ability to predict qCT metrics, quantified by the coefficient of determination\n \n \n (<\/mml:mo>\n \n \n \n R<\/mml:mi>\n <\/mml:mrow>\n 2<\/mml:mn>\n <\/mml:msup>\n <\/mml:mrow>\n )<\/mml:mo>\n <\/mml:math>\n <\/jats:inline-formula>\n .\n <\/jats:p>\n <\/jats:sec>\n \n Results<\/jats:title>\n \n The model achieved a macro-AUC of 89.3% (95% CI: 86.5, 91.8;\n P<\/jats:italic>\n \u2009&lt;\u20090.001) in differentiating the four classes. Post-COVID-19 emerged as a distinct class from asthma and COPD in the t-SNE embedding space, and its embeddings across two visits captured disease improvement. Additionally, the learned embeddings showed predictive power for several qCT metrics, particularly the Jacobian (\n \n \n \n \n R<\/mml:mi>\n <\/mml:mrow>\n 2<\/mml:mn>\n <\/mml:msup>\n <\/mml:math>\n <\/jats:inline-formula>\n =0.61).\n <\/jats:p>\n <\/jats:sec>\n \n Conclusions<\/jats:title>\n The proposed model effectively differentiated these three lung diseases and provided meaningful embeddings for phenotype characterization, longitudinal assessment, and qCT metric prediction.<\/jats:p>\n <\/jats:sec>","DOI":"10.3389\/fradi.2026.1824991","type":"journal-article","created":{"date-parts":[[2026,8,19]],"date-time":"2026-08-19T14:18:51Z","timestamp":1787149131000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Discriminative feature learning for multiclass lung disease classification using contrastive learning"],"prefix":"10.3389","volume":"6","author":[{"given":"Hafiza Akter","family":"Munira","sequence":"first","affiliation":[{"name":"IIHR-Hydroscience & Engineering, University of Iowa","place":["Iowa City, IA, United States"]},{"name":"Department of Mechanical Engineering, University of Iowa","place":["Iowa City, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xuan","family":"Zhang","sequence":"additional","affiliation":[{"name":"IIHR-Hydroscience & Engineering, University of Iowa","place":["Iowa City, IA, United States"]},{"name":"Department of Mechanical Engineering, University of Iowa","place":["Iowa City, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Prathish K.","family":"Rajaraman","sequence":"additional","affiliation":[{"name":"IIHR-Hydroscience & Engineering, University of Iowa","place":["Iowa City, IA, United States"]},{"name":"Department of Mechanical Engineering, University of Iowa","place":["Iowa City, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alejandro P.","family":"Comellas","sequence":"additional","affiliation":[{"name":"Department of Internal Medicine, University of Iowa","place":["Iowa City, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Eric A.","family":"Hoffman","sequence":"additional","affiliation":[{"name":"Department of Radiology, University of Iowa","place":["Iowa City, IA, United States"]},{"name":"Roy J. Carver Department of Biomedical Engineering, University of Iowa","place":["Iowa City, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sean B.","family":"Fain","sequence":"additional","affiliation":[{"name":"Department of Radiology, University of Iowa","place":["Iowa City, IA, United States"]},{"name":"Roy J. Carver Department of Biomedical Engineering, University of Iowa","place":["Iowa City, IA, United States"]},{"name":"Department of Electrical and Computer Engineering, University of Iowa","place":["Iowa City, IA, United States"]},{"name":"Department of Health and Human Physiology, University of Iowa","place":["Iowa City, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jiwoong","family":"Choi","sequence":"additional","affiliation":[{"name":"Division of Pulmonary, Critical Care and Sleep Medicine, University of Kansas School of Medicine","place":["Kansas City, KS, United States"]},{"name":"Bioengineering Program, University of Kansas","place":["Lawrence, KS, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mario","family":"Castro","sequence":"additional","affiliation":[{"name":"Division of Pulmonary, Critical Care and Sleep Medicine, University of Kansas School of Medicine","place":["Kansas City, KS, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mark L.","family":"Schiebler","sequence":"additional","affiliation":[{"name":"School of Medicine & Public Health, University of Wisconsin","place":["Madison, WI, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Elliot","family":"Israel","sequence":"additional","affiliation":[{"name":"Brigham & Women's Hospital, Harvard Medical School","place":["Boston, MA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Serpil C.","family":"Erzurum","sequence":"additional","affiliation":[{"name":"Cleveland Clinic","place":["Cleveland, OH, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tianbao","family":"Yang","sequence":"additional","affiliation":[{"name":"Computer Science & Engineering, Texas A&M University","place":["College Station, TX, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ching-Long","family":"Lin","sequence":"additional","affiliation":[{"name":"IIHR-Hydroscience & Engineering, University of Iowa","place":["Iowa City, IA, United States"]},{"name":"Department of Mechanical Engineering, University of Iowa","place":["Iowa City, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,8,19]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"123","DOI":"10.4103\/jnsm.jnsm_164_24","article-title":"Management of chronic respiratory diseases: a retrospective analysis of demographic predictors and treatment outcomes in asthma and chronic obstructive pulmonary disease","volume":"8","author":"Beigh","year":"2025","journal-title":"J Nat Sci Med"},{"key":"B2","doi-asserted-by":"publisher","first-page":"e0329283","DOI":"10.1371\/journal.pone.0329283","article-title":"Global burden of major chronic respiratory diseases among older adults aged 55 and above from 1990 to 2021: changes, challenges, and predictions amid the pandemic","volume":"20","author":"Xu","year":"2025","journal-title":"PLoS One"},{"key":"B3","doi-asserted-by":"publisher","first-page":"e243374","DOI":"10.1148\/radiol.243374","article-title":"Best practice: international multisociety consensus statement for post\u2013COVID-19 residual abnormalities on chest CT scans","volume":"316","author":"Yoon","year":"2025","journal-title":"Radiology"},{"key":"B4","year":"2024"},{"key":"B5","year":"2024"},{"key":"B6","doi-asserted-by":"publisher","first-page":"889","DOI":"10.4187\/respcare.10757","article-title":"The role of pulmonary function testing in the diagnosis and management of COPD","volume":"68","author":"Haynes","year":"2023","journal-title":"Respir Care"},{"key":"B7","doi-asserted-by":"publisher","first-page":"2119","DOI":"10.1513\/AnnalsATS.201605-418OC","article-title":"Accuracy and quality of spirometry in primary care offices","volume":"13","author":"Hegewald","year":"2016","journal-title":"Ann Am Thorac Soc"},{"key":"B8","doi-asserted-by":"publisher","first-page":"1273253","DOI":"10.3389\/fpubh.2023.1273253","article-title":"Medical image analysis using deep learning algorithms","volume":"11","author":"Li","year":"2023","journal-title":"Front Public Health"},{"key":"B9","doi-asserted-by":"publisher","first-page":"E167","DOI":"10.1148\/radiol.2020203511","article-title":"DeepCOVID-XR: an artificial intelligence algorithm to detect COVID-19 on chest radiographs trained and tested on a large US clinical data set","volume":"299","author":"Wehbe","year":"2021","journal-title":"Radiology"},{"key":"B10","doi-asserted-by":"publisher","first-page":"357","DOI":"10.1186\/s12890-022-02144-2","article-title":"Artificial intelligence to differentiate asthma from COPD in medico-administrative databases","volume":"22","author":"Joumaa","year":"2022","journal-title":"BMC Pulm Med"},{"key":"B11","doi-asserted-by":"publisher","first-page":"999263","DOI":"10.3389\/fphys.2022.999263","article-title":"Contrastive learning and subtyping of post-COVID-19 lung computed tomography images","volume":"13","author":"Li","year":"2022","journal-title":"Front Physiol"},{"key":"B12","author":"Rajpurkar","year":"2017"},{"key":"B13","author":"Khosla","year":"2020"},{"key":"B14","doi-asserted-by":"publisher","first-page":"107848","DOI":"10.1016\/j.patcog.2021.107848","article-title":"Multi-task contrastive learning for automatic CT and x-ray diagnosis of COVID-19","volume":"114","author":"Li","year":"2021","journal-title":"Pattern Recogn"},{"key":"B15","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV48922.2021.00346","article-title":"Big self-supervised models advance medical image classification","author":"Azizi","year":"2021"},{"key":"B16","article-title":"A simple framework for contrastive learning of visual representations","author":"Chen","year":"2020"},{"key":"B17","doi-asserted-by":"publisher","first-page":"e230342","DOI":"10.1148\/ryai.230342","article-title":"Improving fairness of automated chest radiograph diagnosis by contrastive learning","volume":"6","author":"Lin","year":"2024","journal-title":"Radiology: Artificial Intelligence"},{"key":"B18","doi-asserted-by":"publisher","first-page":"103062","DOI":"10.1016\/j.media.2023.103062","article-title":"DrasCLR: a self-supervised framework of learning disease-related and anatomy-specific representation for 3D lung CT images","volume":"92","author":"Ke","year":"2024","journal-title":"Med Image Anal"},{"key":"B19","doi-asserted-by":"publisher","first-page":"e190006","DOI":"10.1148\/ryai.2020190006","article-title":"Deep learning\u2013based approach for automated assessment of interstitial lung disease in systemic sclerosis on CT images","volume":"2","author":"Chassagnon","year":"2020","journal-title":"Radiology: Artificial Intelligence"},{"key":"B20","doi-asserted-by":"publisher","first-page":"2777","DOI":"10.1007\/s10489-020-01943-6","article-title":"Deep neural network to detect COVID-19: one architecture for both CT scans and chest x-rays","volume":"51","author":"Mukherjee","year":"2021","journal-title":"Appl Intell"},{"key":"B21","doi-asserted-by":"publisher","first-page":"e240680","DOI":"10.1148\/ryai.240680","article-title":"Single inspiratory chest CT\u2013based generative deep learning models to evaluate functional small airways disease","volume":"7","author":"Zhang","year":"2025","journal-title":"Radiol Artif Intell"},{"key":"B22","doi-asserted-by":"publisher","first-page":"2252010","DOI":"10.1142\/S0218001422520103","article-title":"Deep learning for COVID-19 screening using chest x-rays in 2020: a systematic review","volume":"36","author":"Santosh","year":"2022","journal-title":"Int J Pattern Recognit Artif Intell"},{"key":"B23","doi-asserted-by":"publisher","first-page":"14","DOI":"10.1016\/j.jaci.2006.10.025","article-title":"Severe asthma: lessons from the severe asthma research program","volume":"119","author":"Wenzel","year":"2007","journal-title":"J Allergy Clin Immunol"},{"key":"B24","doi-asserted-by":"publisher","first-page":"32","DOI":"10.3109\/15412550903499522","article-title":"Genetic epidemiology of COPD (COPDGene) study design","volume":"7","author":"Regan","year":"2011","journal-title":"COPD"},{"key":"B25","doi-asserted-by":"publisher","first-page":"1286","DOI":"10.1152\/japplphysiol.01094.2014","article-title":"Quantitative assessment of multiscale structural and functional alterations in asthmatic populations","volume":"118","author":"Choi","year":"2015","journal-title":"J Appl Physiol"},{"key":"B26","doi-asserted-by":"publisher","first-page":"2489","DOI":"10.1016\/j.jbiomech.2011.06.009","article-title":"Three-dimensional characterization of regional lung deformation","volume":"44","author":"Amelon","year":"2011","journal-title":"J Biomech"},{"key":"B27","doi-asserted-by":"publisher","first-page":"690","DOI":"10.1016\/j.jaci.2016.11.053","article-title":"Quantitative computed tomographic imaging\u2013based clustering differentiates asthmatic subgroups with distinctive clinical phenotypes","volume":"140","author":"Choi","year":"2017","journal-title":"J Allergy Clin Immunol"},{"key":"B28","article-title":"Condconv: conditionally parameterized convolutions for efficient inference","volume":"32","author":"Yang","year":"2019","journal-title":"Adv Neural Inf Process Syst"},{"key":"B29","doi-asserted-by":"publisher","DOI":"10.1186\/s12931-026-03784-2","article-title":"Bronchovascular texture pattern on quantitative CT reveals airway and vascular remodeling in post-COVID-19 lungs","author":"Zhang","year":"2026","journal-title":"Respir Res"},{"key":"B30","doi-asserted-by":"publisher","first-page":"178","DOI":"10.1186\/s12931-018-0888-7","article-title":"Imaging-based clusters in current smokers of the COPD cohort associate with clinical characteristics: the SubPopulations and intermediate outcome measures in COPD study (SPIROMICS)","volume":"19","author":"Haghighi","year":"2018","journal-title":"Respir Res"},{"key":"B31","doi-asserted-by":"publisher","first-page":"153","DOI":"10.1186\/s12931-019-1121-z","article-title":"Imaging-based clusters in former smokers of the COPD cohort associate with clinical characteristics: the SubPopulations and intermediate outcome measures in COPD study (SPIROMICS)","volume":"20","author":"Haghighi","year":"2019","journal-title":"Respir Res"},{"key":"B32","doi-asserted-by":"publisher","first-page":"107093","DOI":"10.1016\/j.ejps.2025.107093","article-title":"Assessment of ventilation heterogeneity and particle deposition in asthmatics using combined SPECT\/CT imaging and computational modeling approaches","volume":"209","author":"Zhang","year":"2025","journal-title":"Eur J Pharm Sci"},{"key":"B33","doi-asserted-by":"publisher","first-page":"1477","DOI":"10.2147\/COPD.S301466","article-title":"Longitudinal imaging-based clusters in former smokers of the COPD cohort associate with clinical characteristics: the subpopulations and intermediate outcome measures in COPD study (spiromics)","author":"Zou","year":"2021","journal-title":"Int J Chron Obstruct Pulmon Dis"},{"key":"B34","doi-asserted-by":"publisher","first-page":"4916","DOI":"10.1038\/s41598-021-84547-5","article-title":"Latent traits of lung tissue patterns in former smokers derived by dual channel deep learning in computed tomography images","volume":"11","author":"Li","year":"2021","journal-title":"Sci Rep"},{"key":"B35","doi-asserted-by":"publisher","first-page":"3523","DOI":"10.2147\/COPD.S149382","article-title":"Smoking history and emphysema in asthma\u2013COPD overlap","author":"Kurashima","year":"2017","journal-title":"Int J Chron Obstruct Pulmon Dis"},{"key":"B36","doi-asserted-by":"crossref","first-page":"335","DOI":"10.1007\/3-540-28397-8_36","article-title":"klaR Analyzing German Business Cycles. In: Baier D, Decker R, Schmidt-Thieme L, editors","volume-title":"Data Analysis and Decision Support","author":"Weihs","year":"2005"}],"container-title":["Frontiers in Radiology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fradi.2026.1824991\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,19]],"date-time":"2026-08-19T14:18:53Z","timestamp":1787149133000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fradi.2026.1824991\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,19]]},"references-count":36,"alternative-id":["10.3389\/fradi.2026.1824991"],"URL":"https:\/\/doi.org\/10.3389\/fradi.2026.1824991","relation":{},"ISSN":["2673-8740"],"issn-type":[{"value":"2673-8740","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,19]]},"article-number":"1824991"}}