A Lung CT Foundation Model Facilitating Disease Diagnosis and Medical Imaging

A Lung CT Foundation Model Facilitating Disease Diagnosis and Medical Imaging | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article A Lung CT Foundation Model Facilitating Disease Diagnosis and Medical Imaging Yuchen Guo, Zebin Gao, Guoxun Zhang, Hengrui Liang, Jiaxin Liu, and 10 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5262017/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 03 Dec, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract The concomitant development and evolution of lung computed tomography (CT) and artificial intelligence (AI) has allowed non-invasive lung imaging to be a key part of the clinical care of patients with major diseases, such as lung cancer. However, the paucity of labeled lung CT data has limited the training highly efficacious AI models and thereby has retarded broad-scale adoption and deployment of AI-based lung CT imaging in the real-world clinical setting. In this paper, We introduce LCTfound, a foundational model that encodes images along with correlated clinical information, into a neural network. LCTfound used self-supervised learning pre-trained by diffusion models using a large dataset containing 105,184 lung CT scans (totaling more than 28 million images) from multiple centers. LCTfound was evaluated on 8 categories of lung CT tasks, ranging from scanning-level clinical diagnosis to pixel-level image restoration, including segmentation of mediastinal neoplasm, diagnosis of pulmonary alveolar proteinosis, prognosis of non-small cell lung cancer, prediction of major pathological response to neoadjuvant chemoimmunotherapy, whole lung 3D modeling for surgical navigation, virtual lung computed tomography angiography(CTA), reconstruction of lung CT from sparse views, and enhancement of low-dose CT images. Equipped with the robust few-shot learning capability, LCTfound outperformed the previously state-of-the-art pre-trained models in all the above tasks. LCTfound is a major advancements in self-supervised representation learning on lung CT, laying the groundwork for a foundational model that operates with high efficacy across the spectrum of low-level and high-level clinical tasks and serving a dual purpose in aiding in clinical diagnosis of lung diseases and improving the quality of lung CT imaging Health sciences/Biomarkers/Diagnostic markers Health sciences/Biomarkers/Prognostic markers Biological sciences/Biological techniques/Imaging Full Text Additional Declarations There is NO Competing Interest. Supplementary Files LCTfoundsupplementarymaterials.pdf Supplementary materials LCTfoundsupplementaryvideo1.mp4 The 3D visualization of mediastinal neoplasms segmentation results LCTfoundsupplementaryvideo2.mp4 The role of the whole lung segmentation model in pulmonary nodule resection surgery LCTfoundsupplementaryvideo3.mp4 The 3D visualization of whole lung segmentation results Cite Share Download PDF Status: Published Journal Publication published 03 Dec, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5262017","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":408266609,"identity":"8b039fbc-af10-4668-8530-6d8bf65e3c9f","order_by":0,"name":"Yuchen Guo","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA50lEQVRIiWNgGAWjYBACgwMMbAjeByBmYydFC+MMkBZmAlokG5C0MPOASQJa+CWSnz34uIMhsYF/jeFjm1/b5PmYGRg/fMzBrYVNIs3ccOYZoBaJN8bGuX23DduYGZglZ27DpyWHTZq3jSG3QeKMmXRuz21GoBY2Zl5CWv7CtFj23LYnTgsjSAt/j5k0w4/biYS18Dwzk+xtk6hvk2ArNuxtuJ3cxszYjN8v7MnPJH622Rjz8x/e+ODHn9u289ubD374iEcLFEgAXZgAjMs2EIexgaB6COA/ACT+EKl4FIyCUTAKRhQAAMMLRsbqGpEOAAAAAElFTkSuQmCC","orcid":"","institution":"Tsinghua University","correspondingAuthor":true,"prefix":"","firstName":"Yuchen","middleName":"","lastName":"Guo","suffix":""},{"id":408266610,"identity":"bc759514-836f-4ffc-9891-160a5dd647ad","order_by":1,"name":"Zebin Gao","email":"","orcid":"","institution":"Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Zebin","middleName":"","lastName":"Gao","suffix":""},{"id":408266611,"identity":"fbf304a2-8b3e-4ff3-bb5b-54428483c3be","order_by":2,"name":"Guoxun Zhang","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Guoxun","middleName":"","lastName":"Zhang","suffix":""},{"id":408266612,"identity":"a8ade915-642a-433a-8197-e79732d7710e","order_by":3,"name":"Hengrui Liang","email":"","orcid":"https://orcid.org/0000-0002-1155-5963","institution":"Department of Thoracic Surgery and Oncology, the First Affiliated Hospital of Guangzhou Medical University","correspondingAuthor":false,"prefix":"","firstName":"Hengrui","middleName":"","lastName":"Liang","suffix":""},{"id":408266613,"identity":"561cab0c-a645-4fdf-a90e-df1b12ae4e77","order_by":4,"name":"Jiaxin Liu","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Jiaxin","middleName":"","lastName":"Liu","suffix":""},{"id":408266614,"identity":"edd5d61e-b223-4aa7-897e-bcd1d9a7ae60","order_by":5,"name":"Liangdi Ma","email":"","orcid":"","institution":"School of software and BNRist, Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Liangdi","middleName":"","lastName":"Ma","suffix":""},{"id":408266615,"identity":"a1bad3d6-fb8d-4ba1-8a85-7da5661caebe","order_by":6,"name":"Tianyun Wang","email":"","orcid":"","institution":"Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Tianyun","middleName":"","lastName":"Wang","suffix":""},{"id":408266616,"identity":"1758b331-487c-4a21-b876-cb5c98c9b7f7","order_by":7,"name":"Yanchen Guo","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Yanchen","middleName":"","lastName":"Guo","suffix":""},{"id":408266617,"identity":"1e1594d7-186d-46b1-a92e-82eded9c0132","order_by":8,"name":"Yujia Chen","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Yujia","middleName":"","lastName":"Chen","suffix":""},{"id":408266618,"identity":"033244e1-00b4-4fcb-9a86-8033159b202f","order_by":9,"name":"Yan Zeping","email":"","orcid":"","institution":"Department of Thoracic Oncology and Surgery, China State Key Laboratory of Respiratory Disease \u0026 National Clinical Research Center for Respiratory Disease, the First Affiliated Hospital of Guangzhou","correspondingAuthor":false,"prefix":"","firstName":"Yan","middleName":"","lastName":"Zeping","suffix":""},{"id":408266619,"identity":"a08e1367-8e65-46b5-b05f-1cf2f8becefa","order_by":10,"name":"Xiangru Chen","email":"","orcid":"","institution":"Hangzhou Zhuoxi Institute of Brain and Intelligence","correspondingAuthor":false,"prefix":"","firstName":"Xiangru","middleName":"","lastName":"Chen","suffix":""},{"id":408266620,"identity":"afe8aa51-29a4-4d0a-a9f1-1246f834eb79","order_by":11,"name":"Jianxing He","email":"","orcid":"https://orcid.org/0000-0003-1737-8192","institution":"Department of Thoracic Surgery and Oncology, the First Affiliated Hospital of Guangzhou Medical University","correspondingAuthor":false,"prefix":"","firstName":"Jianxing","middleName":"","lastName":"He","suffix":""},{"id":408266621,"identity":"5ba9bb4e-22ca-428b-ad9c-dec7a0165d56","order_by":12,"name":"Feng Xu","email":"","orcid":"https://orcid.org/0000-0002-0953-1057","institution":"School of software and BNRist, Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Feng","middleName":"","lastName":"Xu","suffix":""},{"id":408266622,"identity":"c0f26f17-ed93-482e-9e0d-2a1701389e8a","order_by":13,"name":"Tien Yin Wong","email":"","orcid":"https://orcid.org/0000-0002-8448-1264","institution":"Tsinghua Medicine, Beijing Tsinghua Changgung Hospital, Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Tien","middleName":"Yin","lastName":"Wong","suffix":""},{"id":408266623,"identity":"c69dd9d7-22c1-42a3-a2f7-3608d7b920d1","order_by":14,"name":"Qionghai Dai","email":"","orcid":"https://orcid.org/0000-0001-7043-3061","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Qionghai","middleName":"","lastName":"Dai","suffix":""}],"badges":[],"createdAt":"2024-10-14 14:43:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5262017/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5262017/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-025-66620-z","type":"published","date":"2025-12-03T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":99419426,"identity":"e5c33ef7-fb63-4f57-a730-dc6271ccc2e6","added_by":"auto","created_at":"2026-01-03 08:09:10","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3361169,"visible":true,"origin":"","legend":"Article File","description":"","filename":"LCTfoundmain.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5262017/v1_covered_5417b4f6-9f22-48d8-9d66-404be95f14f5.pdf"},{"id":74986531,"identity":"4afcf579-1604-4872-a7c3-70f48b971067","added_by":"auto","created_at":"2025-01-29 06:18:13","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":33282595,"visible":true,"origin":"","legend":"Supplementary materials","description":"","filename":"LCTfoundsupplementarymaterials.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5262017/v1/40e44dbfdc431cc1874aa607.pdf"},{"id":74986533,"identity":"c5f08316-3457-4cc4-bcdc-ad7a213e37e9","added_by":"auto","created_at":"2025-01-29 06:18:14","extension":"mp4","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":50546286,"visible":true,"origin":"","legend":"\u003cp\u003eThe 3D visualization of mediastinal neoplasms segmentation results\u003c/p\u003e","description":"","filename":"LCTfoundsupplementaryvideo1.mp4","url":"https://assets-eu.researchsquare.com/files/rs-5262017/v1/7dd05a0711da0f274ac4fb9c.mp4"},{"id":74986529,"identity":"2608e15f-b7b1-490f-8105-7152f5a836f5","added_by":"auto","created_at":"2025-01-29 06:18:13","extension":"mp4","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":2061310,"visible":true,"origin":"","legend":"\u003cp\u003eThe role of the whole lung segmentation model in pulmonary nodule resection surgery\u003c/p\u003e","description":"","filename":"LCTfoundsupplementaryvideo2.mp4","url":"https://assets-eu.researchsquare.com/files/rs-5262017/v1/9c21028c9871abeeca403c02.mp4"},{"id":74988001,"identity":"2bd6f25f-6760-4fa5-9be1-66dff2d06e2f","added_by":"auto","created_at":"2025-01-29 06:26:14","extension":"mp4","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":61283490,"visible":true,"origin":"","legend":"The 3D visualization of whole lung segmentation results","description":"","filename":"LCTfoundsupplementaryvideo3.mp4","url":"https://assets-eu.researchsquare.com/files/rs-5262017/v1/02b80092dcb0c22e5906480d.mp4"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"A Lung CT Foundation Model Facilitating Disease Diagnosis and Medical Imaging","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5262017/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5262017/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"The concomitant development and evolution of lung computed tomography (CT) and artificial intelligence (AI) has allowed non-invasive lung imaging to be a key part of the clinical care of patients with major diseases, such as lung cancer. However, the paucity of labeled lung CT data has limited the training highly efficacious AI models and thereby has retarded broad-scale adoption and deployment of AI-based lung CT imaging in the real-world clinical setting. In this paper, We introduce LCTfound, a foundational model that encodes images along with correlated clinical information, into a neural network. LCTfound used self-supervised learning pre-trained by diffusion models using a large dataset containing 105,184 lung CT scans (totaling more than 28 million images) from multiple centers. LCTfound was evaluated on 8 categories of lung CT tasks, ranging from scanning-level clinical diagnosis to pixel-level image restoration, including segmentation of mediastinal neoplasm, diagnosis of pulmonary alveolar proteinosis, prognosis of non-small cell lung cancer, prediction of major pathological response to neoadjuvant chemoimmunotherapy, whole lung 3D modeling for surgical navigation, virtual lung computed tomography angiography(CTA), reconstruction of lung CT from sparse views, and enhancement of low-dose CT images. Equipped with the robust few-shot learning capability, LCTfound outperformed the previously state-of-the-art pre-trained models in all the above tasks. LCTfound is a major advancements in self-supervised representation learning on lung CT, laying the groundwork for a foundational model that operates with high efficacy across the spectrum of low-level and high-level clinical tasks and serving a dual purpose in aiding in clinical diagnosis of lung diseases and improving the quality of lung CT imaging","manuscriptTitle":"A Lung CT Foundation Model Facilitating Disease Diagnosis and Medical Imaging","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-29 06:18:08","doi":"10.21203/rs.3.rs-5262017/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"600baa7f-e629-4eb0-98e6-00962cd6a57a","owner":[],"postedDate":"January 29th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":43529495,"name":"Health sciences/Biomarkers/Diagnostic markers"},{"id":43529496,"name":"Health sciences/Biomarkers/Prognostic markers"},{"id":43529497,"name":"Biological sciences/Biological techniques/Imaging"}],"tags":[],"updatedAt":"2026-01-03T08:09:03+00:00","versionOfRecord":{"articleIdentity":"rs-5262017","link":"https://doi.org/10.1038/s41467-025-66620-z","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2025-12-03 05:00:00","publishedOnDateReadable":"December 3rd, 2025"},"versionCreatedAt":"2025-01-29 06:18:08","video":"","vorDoi":"10.1038/s41467-025-66620-z","vorDoiUrl":"https://doi.org/10.1038/s41467-025-66620-z","workflowStages":[]},"version":"v1","identity":"rs-5262017","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5262017","identity":"rs-5262017","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}