{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T06:15:19Z","timestamp":1788243319667,"version":"build-2803163510"},"reference-count":27,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"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. Med."],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>We aimed to construct and validate an intelligent differential diagnosis model that integrates U-Net-based automatic segmentation with a deep learning classification model, and to evaluate its diagnostic performance and clinical value for distinguishing tuberculous pleural effusion (TPE) from malignant pleural effusion (MPE).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>A total of 281 patients with pleural effusion confirmed by etiological or pathological evidence between January 2018 and August 2025 were included, comprising 143 patients with TPE and 138 with MPE. First, a U-Net model was employed to automatically segment pleural lesion regions on chest computed tomography (CT) images and extract regions of interest (ROIs). Subsequently, based on the segmentation results, a radiomics model, a two-dimensional deep learning (DL2D) model, and a comprehensive model integrating clinical features were constructed. Multiple machine learning algorithms, including support vector machines (SVMs), random forests (RFs), and extremely randomized trees (ERTs), were utilized for model construction and comparison. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC), sensitivity, specificity, calibration curves, decision curve analysis (DCA), and the integrated discrimination improvement (IDI) and net reclassification improvement (NRI) indices.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      The U-Net segmentation model achieved Dice coefficients of 0.873 and 0.862 in the training and test sets, respectively, indicating good segmentation performance. In the test set, the comprehensive model demonstrated the best performance, with an AUC of 0.934 (95% CI 0.8733\u20130.9955), sensitivity of 0.875, and specificity of 0.900. Its performance was superior to that of the clinical model (AUC\u202f=\u202f0.767), the radiomics model (AUC\u202f=\u202f0.841), and the DL2D model (AUC\u202f=\u202f0.776). DCA confirmed that the comprehensive model provided a higher net clinical benefit across a wide range of threshold probabilities. Furthermore, IDI and NRI analyses indicated that the comprehensive model significantly improved predictive performance relative to the individual models (\n                      <jats:italic>p<\/jats:italic>\n                      \u202f&amp;lt;\u202f0.05).\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>The model combining U-Net-based automatic segmentation with a deep learning classification model exhibited excellent and balanced diagnostic performance for differentiating TPE from MPE. It has the potential to provide an objective, stable, and scalable intelligent decision-support tool for clinical practice.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fmed.2026.1855938","type":"journal-article","created":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T05:35:55Z","timestamp":1788240955000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Differential diagnosis model for tuberculous and malignant pleural effusion combining U-Net automatic segmentation and deep learning"],"prefix":"10.3389","volume":"13","author":[{"given":"Chang","family":"Song","sequence":"first","affiliation":[{"name":"Department of Tuberculosis, The Fourth People\u2019s Hospital of Nanning","place":["Nanning, Guangxi, China"]},{"name":"Clinical Medical School, Guangxi Medical University","place":["Nanning, Guangxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chun-Yan","family":"Zhao","sequence":"additional","affiliation":[{"name":"Department of Tuberculosis, The Fourth People\u2019s Hospital of Nanning","place":["Nanning, Guangxi, China"]},{"name":"Clinical Medical School, Guangxi Medical University","place":["Nanning, Guangxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shu-Lin","family":"Song","sequence":"additional","affiliation":[{"name":"Department of Radiology, The Fourth People\u2019s Hospital of Nanning","place":["Nanning, Guangxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xue-Wen","family":"Huang","sequence":"additional","affiliation":[{"name":"Department of Tuberculosis, The Fourth People\u2019s Hospital of Nanning","place":["Nanning, Guangxi, China"]},{"name":"Clinical Medical School, Guangxi Medical University","place":["Nanning, Guangxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hang-Biao","family":"Qiang","sequence":"additional","affiliation":[{"name":"Department of Tuberculosis, The Fourth People\u2019s Hospital of Nanning","place":["Nanning, Guangxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiao-Shi","family":"Lin","sequence":"additional","affiliation":[{"name":"Clinical Medical School, Guangxi Medical University","place":["Nanning, Guangxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhen-Tao","family":"Huang","sequence":"additional","affiliation":[{"name":"Clinical Medical School, Guangxi Medical University","place":["Nanning, Guangxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhou-Hua","family":"Xie","sequence":"additional","affiliation":[{"name":"Department of Tuberculosis, The Fourth People\u2019s Hospital of Nanning","place":["Nanning, Guangxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Qing-Dong","family":"Zhu","sequence":"additional","affiliation":[{"name":"Department of Tuberculosis, The Fourth People\u2019s Hospital of Nanning","place":["Nanning, Guangxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,9,1]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"189","DOI":"10.1183\/16000617.0019-2016","article-title":"Malignant pleural effusion: from bench to bedside","volume":"25","author":"Psallidas","year":"2016","journal-title":"Eur Respir Rev"},{"key":"ref2","doi-asserted-by":"publisher","first-page":"962","DOI":"10.1111\/resp.13673","article-title":"Tuberculous pleural effusion","volume":"24","author":"Shaw","year":"2019","journal-title":"Respirology"},{"key":"ref3","doi-asserted-by":"publisher","first-page":"680","DOI":"10.1186\/s12967-023-04492-x","article-title":"Simultaneous diagnosis of tuberculous pleurisy and malignant pleural effusion using metagenomic next-generation sequencing (mNGS)","volume":"21","author":"Xu","year":"2023","journal-title":"J Transl Med"},{"key":"ref4","doi-asserted-by":"publisher","first-page":"363","DOI":"10.1159\/000490732","article-title":"Sensitivity of initial thoracentesis for malignant pleural effusion stratified by tumor type in patients with strong evidence of metastatic disease","volume":"96","author":"Grosu","year":"2018","journal-title":"Respiration"},{"key":"ref5","doi-asserted-by":"publisher","first-page":"161","DOI":"10.1016\/j.arbres.2013.11.007","article-title":"Etiology of pleural effusions: analysis of more than 3,000 consecutive thoracenteses","volume":"50","author":"Porcel","year":"2014","journal-title":"Arch Bronconeumol"},{"key":"ref6","doi-asserted-by":"publisher","first-page":"62","DOI":"10.1159\/000468545","article-title":"Diagnostic accuracy of combinations of tumor markers for malignant pleural effusion: an updated Meta-analysis","volume":"94","author":"Yang","year":"2017","journal-title":"Respiration"},{"key":"ref7","doi-asserted-by":"publisher","first-page":"311","DOI":"10.1159\/000356764","article-title":"Prognostic factors in patients presenting with pleural effusion revealing malignancy","volume":"87","author":"Anevlavis","year":"2014","journal-title":"Respiration"},{"key":"ref8","doi-asserted-by":"publisher","first-page":"475","DOI":"10.1016\/j.mcna.2018.12.007","article-title":"Update in the management of pleural effusions","volume":"103","author":"Aboudara","year":"2019","journal-title":"Med Clin North Am"},{"key":"ref9","doi-asserted-by":"publisher","first-page":"154760","DOI":"10.1016\/j.jcrc.2024.154760","article-title":"Assessment of the comparative agreement between chest radiographs and CT scans in intensive care units","volume":"82","author":"Brooks","year":"2024","journal-title":"J Crit Care"},{"key":"ref10","doi-asserted-by":"publisher","first-page":"1911","DOI":"10.3390\/life13091911","article-title":"Performance of deep-learning solutions on lung nodule malignancy classification: a systematic review","volume":"13","author":"Liang","year":"2023","journal-title":"Life"},{"key":"ref11","doi-asserted-by":"publisher","first-page":"56","DOI":"10.1007\/s11547-023-01730-6","article-title":"Combined model integrating deep learning, radiomics, and clinical data to classify lung nodules at chest CT","volume":"129","author":"Lin","year":"2024","journal-title":"Radiol Med"},{"key":"ref12","doi-asserted-by":"publisher","first-page":"748","DOI":"10.3390\/diagnostics13101748","article-title":"Acoustic-based deep learning architectures for lung disease diagnosis: a comprehensive overview","volume":"13","author":"Sfayyih","year":"2023","journal-title":"Diagnostics"},{"key":"ref13","doi-asserted-by":"publisher","first-page":"2987","DOI":"10.1038\/s41598-023-29656-z","article-title":"Deep learning ensemble 2D CNN approach towards the detection of lung cancer","volume":"13","author":"Shah","year":"2023","journal-title":"Sci Rep"},{"key":"ref14","doi-asserted-by":"publisher","first-page":"100416","DOI":"10.1016\/j.health.2025.100416","article-title":"An analytics-driven review of U-net for medical image segmentation","volume":"8","author":"Neha","year":"2025","journal-title":"Healthc Anal"},{"key":"ref15","doi-asserted-by":"publisher","first-page":"102800","DOI":"10.1016\/j.artmed.2024.102800","article-title":"CSCA U-net: a channel and space compound attention CNN for medical image segmentation","volume":"150","author":"Shu","year":"2024","journal-title":"Artif Intell Med"},{"key":"ref16","doi-asserted-by":"publisher","first-page":"383","DOI":"10.5624\/isd.20220105","article-title":"Comparison of multi-label U-net and mask R-CNN for panoramic radiograph segmentation to detect periodontitis","volume":"52","author":"Widyaningrum","year":"2022","journal-title":"Imaging Sci Dent"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"2451","DOI":"10.1128\/jcm.00823-15","article-title":"Interferon gamma release assays for diagnosis of pleural tuberculosis: a systematic review and Meta-analysis","volume":"53","author":"Aggarwal","year":"2015","journal-title":"J Clin Microbiol"},{"key":"ref18","doi-asserted-by":"publisher","first-page":"880","DOI":"10.1378\/chest.06-2063","article-title":"Diagnosis and treatment of tuberculous pleural effusion in 2006","volume":"131","author":"Gopi","year":"2007","journal-title":"Chest"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"2743","DOI":"10.1038\/s41598-022-06685-8","article-title":"A meta-analysis of Th1 and Th2 cytokine profiles differentiating tuberculous from malignant pleural effusion","volume":"12","author":"Zeng","year":"2022","journal-title":"Sci Rep"},{"key":"ref20","doi-asserted-by":"publisher","first-page":"e0329668","DOI":"10.1371\/journal.pone.0329668","article-title":"Machine learning for predicting the diagnosis of tuberculous versus malignant pleural effusion: external validation and accuracy in two different settings","volume":"20","author":"Garcia-Zamalloa","year":"2025","journal-title":"PLoS One"},{"key":"ref21","doi-asserted-by":"publisher","first-page":"110655","DOI":"10.1016\/j.clinbiochem.2023.110655","article-title":"Construction of a multi-classified decision tree model for identifying malignant pleural effusion and tuberculous pleural effusion","volume":"120","author":"Liu","year":"2023","journal-title":"Clin Biochem"},{"key":"ref22","doi-asserted-by":"publisher","first-page":"1419343","DOI":"10.3389\/fonc.2024.1419343","article-title":"The value of computed tomography-based radiomics for predicting malignant pleural effusions","volume":"14","author":"Xing","year":"2024","journal-title":"Front Oncol"},{"key":"ref23","doi-asserted-by":"publisher","first-page":"e18056","DOI":"10.1016\/j.heliyon.2023.e18056","article-title":"Novel clinical radiomic nomogram method for differentiating malignant from non-malignant pleural effusions","volume":"9","author":"Han","year":"2023","journal-title":"Heliyon"},{"key":"ref24","doi-asserted-by":"publisher","first-page":"213","DOI":"10.1007\/s12672-023-00835-8","article-title":"Development and validation of a radiomics nomogram for diagnosis of malignant pleural effusion","volume":"14","author":"Wei","year":"2023","journal-title":"Discov Oncol"},{"key":"ref25","doi-asserted-by":"publisher","first-page":"435","DOI":"10.1016\/j.arbres.2013.07.006","article-title":"Tuberculous pleural effusion","volume":"50","author":"Ferreiro","year":"2014","journal-title":"Arch Bronconeumol"},{"key":"ref26","doi-asserted-by":"publisher","first-page":"193","DOI":"10.1053\/crad.2000.0573","article-title":"Thoracic computed tomography in patients with suspected malignant pleural effusions","volume":"56","author":"Traill","year":"2001","journal-title":"Clin Radiol"},{"key":"ref27","doi-asserted-by":"publisher","first-page":"17","DOI":"10.4081\/monaldi.2005.653","article-title":"CT in differential diagnosis of benign and malignant pleural disease","volume":"63","author":"Yilmaz","year":"2005","journal-title":"Monaldi Arch Chest Dis"}],"container-title":["Frontiers in Medicine"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2026.1855938\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T05:35:56Z","timestamp":1788240956000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2026.1855938\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,1]]},"references-count":27,"alternative-id":["10.3389\/fmed.2026.1855938"],"URL":"https:\/\/doi.org\/10.3389\/fmed.2026.1855938","relation":{},"ISSN":["2296-858X"],"issn-type":[{"value":"2296-858X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,1]]},"article-number":"1855938"}}