Delta-Enhanced CT-Based 2.5D Deep Learning Model for Noninvasive Prediction of Leptomeningeal Metastasis in Lung Adenocarcinoma

Delta-Enhanced CT-Based 2.5D Deep Learning Model for Noninvasive Prediction of Leptomeningeal Metastasis in Lung Adenocarcinoma | 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 Research Article Delta-Enhanced CT-Based 2.5D Deep Learning Model for Noninvasive Prediction of Leptomeningeal Metastasis in Lung Adenocarcinoma Jiachun Sun, Renke Yu, Jingxiang Su, Bo Sun, Yitao Fan, Yingda Xie, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9115524/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Leptomeningeal metastasis (LM) is a devastating complication of lung adenocarcinoma, the current diagnosis of which relies on invasive cerebrospinal fluid cytology or costly MRI, both of which are limited in sensitivity and accessibility. To enable early, noninvasive LM risk prediction, we developed a 2.5D delta-radiomics deep learning framework based on routinely acquired contrast-enhanced CT scans. The proposed architecture integrates multi-instance learning to aggregate slice-level features into patient-level representations while preserving spatial context and controlling computational complexity. Delta features, derived by comparing pre-LM and post-LM scans, quantified subtle longitudinal changes within intratumoral and peritumoral regions. Multiple classifiers were evaluated, with XGBoost selected for its optimal performance, and Grad-CAM visualization was employed to assess spatial attention, revealing biologically plausible regions that drove model predictions. A nomogram was constructed to facilitate individualized clinical application, and decision curve analysis (DCA) demonstrated a consistent net benefit across a broad range of decision thresholds. In the test set, the delta-based multi-instance learning signature (MILDelta) achieved an AUC of 0.871, outperforming models trained solely on pre-LM (AUC 0.574) or post-LM (AUC 0.661) images, whereas a combined model incorporating imaging signatures and clinical variables, including EGFR mutation status, yielded the highest predictive accuracy, with an AUC of 0.910. This interpretable delta-informed model leverages standard contrast-enhanced CT to achieve individualized, noninvasive LM risk stratification with demonstrated generalizability and clinical utility, offering a scalable tool for timely intervention in precision oncology workflows. leptomeningeal metastasis lung adenocarcinoma delta radiomics 2.5D deep learning model multi-instance learning Full Text Additional Declarations The authors declare no competing interests. Supplementary Files supplementary.docx Cite Share Download PDF Status: Posted 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-9115524","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":605706822,"identity":"4b314d31-72b2-46a3-b982-68149773dc49","order_by":0,"name":"Jiachun Sun","email":"","orcid":"","institution":"Henan Key Laboratory of Cancer Epigenetics, Cancer Institute, The First Affiliated Hospital, College of Clinical Medicine, Henan University of Science and Technology, Luoyang, Henan, China.","correspondingAuthor":false,"prefix":"","firstName":"Jiachun","middleName":"","lastName":"Sun","suffix":""},{"id":605710593,"identity":"9bffd086-c0fb-4c4d-8c95-b058fcdef267","order_by":1,"name":"Renke Yu","email":"","orcid":"","institution":"Department of General Practice, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China.","correspondingAuthor":false,"prefix":"","firstName":"Renke","middleName":"","lastName":"Yu","suffix":""},{"id":605710594,"identity":"bcc9a175-913a-427f-b70a-64713a0efafc","order_by":2,"name":"Jingxiang Su","email":"","orcid":"","institution":"Department of Radiation Oncology, Tianjin Medical University Cancer Institute \u0026 Hospital, Tianjin, China.","correspondingAuthor":false,"prefix":"","firstName":"Jingxiang","middleName":"","lastName":"Su","suffix":""},{"id":605710595,"identity":"2cc10f94-8694-4a84-9c04-fb83486bdf72","order_by":3,"name":"Bo Sun","email":"","orcid":"","institution":"Henan Key Laboratory of Cancer Epigenetics, Cancer Institute, The First Affiliated Hospital, College of Clinical Medicine, Henan University of Science and Technology, Luoyang, Henan, China.","correspondingAuthor":false,"prefix":"","firstName":"Bo","middleName":"","lastName":"Sun","suffix":""},{"id":605710596,"identity":"0b8a4d6d-9593-4029-b7a6-16f9a1f2aef5","order_by":4,"name":"Yitao Fan","email":"","orcid":"","institution":"Henan Key Laboratory of Cancer Epigenetics, Cancer Institute, The First Affiliated Hospital, College of Clinical Medicine, Henan University of Science and Technology, Luoyang, Henan, China.","correspondingAuthor":false,"prefix":"","firstName":"Yitao","middleName":"","lastName":"Fan","suffix":""},{"id":605710597,"identity":"f7e0112f-6050-456b-b165-83c8e66671ea","order_by":5,"name":"Yingda Xie","email":"","orcid":"","institution":"Henan Key Laboratory of Cancer Epigenetics, Cancer Institute, The First Affiliated Hospital, College of Clinical Medicine, Henan University of Science and Technology, Luoyang, Henan, China.","correspondingAuthor":false,"prefix":"","firstName":"Yingda","middleName":"","lastName":"Xie","suffix":""},{"id":605710598,"identity":"5507a014-3464-4f26-b96d-125c9e61bb35","order_by":6,"name":"Hanliang Jiang","email":"","orcid":"","institution":"Department of Respiratory and Critical Care Medicine, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China.","correspondingAuthor":false,"prefix":"","firstName":"Hanliang","middleName":"","lastName":"Jiang","suffix":""},{"id":605710599,"identity":"9fb8af89-3c5c-448c-93e9-1637dc242f6f","order_by":7,"name":"Lu Zheng","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5UlEQVRIie3PsYrCQBCA4VkWkmYx7YiCrzCyEIQLvszC2iynpZ0BCxvBNqAP4SMct1yquz6FjY2VRQo5UlhoFCyT2AnuXwxTzFcMgMv1imE5JreVQ07RsCGhO2HJRKvnCBf5N4vrRG89T/cnshAs/qyM6IuDb3+2VYRt0pHsXgn+fiplaNcCoXVWRTiasINXQmCkNXTggCKsJB6O/+8kOEo7IMviOiLQeO28JGikgiYEUYcdoBFgdpT9JWnl1f3SS9ShXUw/IFgZicU5Gga+TStJGRcAs/jxXd15GSuaXLlcLtf7dgFKX0EWqwtbfAAAAABJRU5ErkJggg==","orcid":"","institution":"Department of Medical Oncology, The Affiliated Cancer Hospital of Zhengzhou University and Henan Cancer Hospital, Zhengzhou, Henan, China.","correspondingAuthor":true,"prefix":"","firstName":"Lu","middleName":"","lastName":"Zheng","suffix":""},{"id":605710600,"identity":"e661c175-9c79-4a50-a9da-105d29c052e1","order_by":8,"name":"Weidong Han","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAu0lEQVRIiWNgGAWjYBACAxDBU8HATKqWMyRr4W0jxWHm7GePSbydV8fO336A8XHFLwZ5c0JaLHvy0iTnbmNjljiTwGx4to/BcGcDIYcdyDG7zbuNh9lAgoFNsrGHIcHgACEt598AtcyRIEXLDZAtDQYQLQ0/iNLyxvznnGMJQL8kNhs2NkgYbiDssBxjgzc1dcn87YcPPmz4YyNP0BYYSGZgYGxgYGyTIFI9ENhBqD/E6xgFo2AUjIKRAwCusDr5rDsaaQAAAABJRU5ErkJggg==","orcid":"","institution":"Department of Colorectal Medical Oncology, Zhejiang Cancer Hospital, Chinese Academy of Sciences, Hangzhou, Zhejiang, China.","correspondingAuthor":true,"prefix":"","firstName":"Weidong","middleName":"","lastName":"Han","suffix":""}],"badges":[],"createdAt":"2026-03-13 14:01:58","currentVersionCode":1,"declarations":{"humanSubjects":true,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":true,"humanSubjectConsent":true,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9115524/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9115524/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":104696232,"identity":"495db2cc-602b-4697-a42f-2d213a96e3eb","added_by":"auto","created_at":"2026-03-16 07:28:31","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1934654,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript20260313.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9115524/v1_covered_ed09c378-ea97-4f04-b8a9-b5dc866024f1.pdf"},{"id":104696190,"identity":"982d28c9-c9ea-403f-b5c8-c91b737f7103","added_by":"auto","created_at":"2026-03-16 07:28:22","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":17975,"visible":true,"origin":"","legend":"","description":"","filename":"supplementary.docx","url":"https://assets-eu.researchsquare.com/files/rs-9115524/v1/3bca586a7f55fef1c7f2e14b.docx"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eDelta-Enhanced CT-Based 2.5D Deep Learning Model for Noninvasive Prediction of Leptomeningeal Metastasis in Lung Adenocarcinoma\u003c/strong\u003e\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[{"identity":"505a1ba1-e3ec-418a-8f40-53df9606d920","identifier":"10.13039/501100001809","name":"National Natural Science Foundation of China","awardNumber":" 62176230","order_by":0}],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Zhejiang Cancer Hospital","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"leptomeningeal metastasis, lung adenocarcinoma, delta radiomics, 2.5D deep learning model, multi-instance learning","lastPublishedDoi":"10.21203/rs.3.rs-9115524/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9115524/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eLeptomeningeal metastasis (LM) is a devastating complication of lung adenocarcinoma, the current diagnosis of which relies on invasive cerebrospinal fluid cytology or costly MRI, both of which are limited in sensitivity and accessibility. To enable early, noninvasive LM risk prediction, we developed a 2.5D delta-radiomics deep learning framework based on routinely acquired contrast-enhanced CT scans. The proposed architecture integrates multi-instance learning to aggregate slice-level features into patient-level representations while preserving spatial context and controlling computational complexity. Delta features, derived by comparing pre-LM and post-LM scans, quantified subtle longitudinal changes within intratumoral and peritumoral regions. Multiple classifiers were evaluated, with XGBoost selected for its optimal performance, and Grad-CAM visualization was employed to assess spatial attention, revealing biologically plausible regions that drove model predictions. A nomogram was constructed to facilitate individualized clinical application, and decision curve analysis (DCA) demonstrated a consistent net benefit across a broad range of decision thresholds. In the test set, the delta-based multi-instance learning signature (MILDelta) achieved an AUC of 0.871, outperforming models trained solely on pre-LM (AUC 0.574) or post-LM (AUC 0.661) images, whereas a combined model incorporating imaging signatures and clinical variables, including EGFR mutation status, yielded the highest predictive accuracy, with an AUC of 0.910. This interpretable delta-informed model leverages standard contrast-enhanced CT to achieve individualized, noninvasive LM risk stratification with demonstrated generalizability and clinical utility, offering a scalable tool for timely intervention in precision oncology workflows.\u003c/p\u003e","manuscriptTitle":"Delta-Enhanced CT-Based 2.5D Deep Learning Model for Noninvasive Prediction of Leptomeningeal Metastasis in Lung Adenocarcinoma","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-16 07:25:38","doi":"10.21203/rs.3.rs-9115524/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"dc97b446-6d3a-4714-820a-e179d3de9589","owner":[],"postedDate":"March 16th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-03-16T07:25:39+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-16 07:25:38","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9115524","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9115524","identity":"rs-9115524","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}