Journal article
Discriminative feature learning for multiclass lung disease classification using contrastive learning
Frontiers in radiology, Vol.6, 1824991
08/19/2026
DOI: 10.3389/fradi.2026.1824991
PMCID: PMC13536901
PMID: 42688736
Abstract
Objectives:
To develop a contrastive learning model for lung disease classification using discriminative CT imaging embeddings.
Methods:
A total of 1,187 subjects were included: asthma (n = 315), COPD (n = 355), post-COVID-19 (n = 375), and healthy controls (n = 142). 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 (R2).
Results:
The model achieved a macro-AUC of 89.3% (95% CI: 86.5, 91.8; P < 0.001) 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 (R2=0.61).
Conclusions:
The proposed model effectively differentiated these three lung diseases and provided meaningful embeddings for phenotype characterization, longitudinal assessment, and qCT metric prediction.
Details
- Title: Subtitle
- Discriminative feature learning for multiclass lung disease classification using contrastive learning
- Creators
- Hafiza Akter Munira - University of IowaXuan Zhang - University of IowaPrathish K. Rajaraman - University of IowaAlejandro P. Comellas - University of IowaEric A. Hoffman - University of IowaSean B. Fain - University of IowaJiwoong Choi - University of KansasMario Castro - University of Kansas Medical CenterMark L. Schiebler - University of Wisconsin–MadisonElliot Israel - Brigham and Women's HospitalSerpil C. Erzurum - Cleveland ClinicTianbao Yang - Texas A&M UniversityChing-Long Lin - University of Iowa, Mechanical Engineering
- Resource Type
- Journal article
- Publication Details
- Frontiers in radiology, Vol.6, 1824991
- DOI
- 10.3389/fradi.2026.1824991
- PMID
- 42688736
- PMCID
- PMC13536901
- NLM abbreviation
- Front Radiol
- ISSN
- 2673-8740
- eISSN
- 2673-8740
- Publisher
- Frontiers Media S.A.
- Grant note
- NIH: R01-HL168116, P30 ES005605, ED P116S210005 National Institutes of Health (NIH) Amgen AstraZeneca GlaxoSmithKline
The author(s) declared that financial support was received for this work and/or its publication. This work was supported in part by NIH R01-HL168116, P30 ES005605, and ED P116S210005. This article is the result of funding in whole or in part by the National Institutes of Health (NIH). It is subject to the NIH Public Access Policy. Through acceptance of this federal funding, NIH has been given a right to make this article publicly available in PubMed Central upon the Official Date of Publication, as defined by NIH. Additionally, this work received funding from Amgen, AstraZeneca, Boehringer-Ingelheim, Genentech, GlaxoSmithKline, Sanofi-Genzyme-Regeneron, and TEVA. The funders were not involved in the study design, collection, analysis, interpretation of data, the writing of this article, or the decision to submit it for publication.
- Language
- English
- Date published
- 08/19/2026
- Academic Unit
- Roy J. Carver Department of Biomedical Engineering; Radiology; Electrical and Computer Engineering; Pulmonary, Critical Care, and Occupational Medicine; ICTS; IIHR--Hydroscience and Engineering; Computer Science; Mechanical Engineering; Internal Medicine
- Record Identifier
- 9985220286802771
Metrics
4 Record Views