Multi-Stain Multi-Level Convolutional Network for Multi-Tissue breast cancer image segmentation

Multi-Stain Multi-Level Convolutional Network for Multi-Tissue breast cancer image segmentation | 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 Help Center Sign In Submit a Preprint Cite Share Download PDF Article Multi-Stain Multi-Level Convolutional Network for Multi-Tissue breast cancer image segmentation Purnendu Mishra, Akash Modi, Sumit Jha, Rajiv Kumar, Kiran Aatre, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3908756/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 Digital pathology and microscopy image analysis are widely employed in the segmentation of digitally scanned IHC slides, primarily to identify cancer and pinpoint regions of interest (ROI) indicative of tumor presence. However, current ROI segmentation models are either stain specific or suffer from the issues of stain and scanner variance due to different staining protocols or modalities across multiple labs. Also, tissues like Ductal Carcinoma in Situ (DCIS), ACINIs etc. are often classified as Tumors due to their structural similarities and color compositions. In this paper, we proposed a novel convolutional neural network (CNN) based Multi-class Tissue Segmentation model for histopathology whole-slide Breast slides which classifies tumor and segments other tissue regions such as Ducts, ACINIs, DCIS, Squamous epithelium, Blood Vessels, Necrosis etc. as separate class. Our unique pixel aligned non-linear merge across spatial resolutions, empowers models with both local and global field of view for accurate detection of various classes. Our proposed model is also able separate bad regions such as folds, artifacts, blurry regions, bubbles etc. from tissue regions using multi-level context from different resolutions of WSI. Multi-phase iterative training with context aware augmentation and increasing noise was used to efficiently train multi-stain generic model with partial and noisy annotations from 513 slides. Our training pipeline used 12 million patches generated using context aware augmentations which made our model stain & scanner invariant across data sources. We have achieved an Area Under curve (AUC) of 0.85 for tumor and 0.93 for other class on 35000 test patches distributed across stains, scanners and data sources. To extrapolate stain & scanner invariance, our model was evaluated on 23000 patches which were for a completely new stain (Hematoxylin and Eosin) from a completely new scanner (Motic) from a different lab. The mean IOU was 0.72 which is on par with model performance on other data sources and scanners. Our results illustrate that the proposed model exhibits remarkable accuracy with consistent performance and broad utility especially in facilitating pathologists’ assessment of breast cancer. Biological sciences/Computational biology and bioinformatics/Machine learning Biological sciences/Computational biology and bioinformatics/Image processing Biological sciences/Cancer/Breast cancer Full Text Additional Declarations (Not answered) 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 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-3908756","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":269926829,"identity":"0469d535-294c-4b05-aa7e-0985bf3c8964","order_by":0,"name":"Purnendu Mishra","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA40lEQVRIiWNgGAWjYPACZiBmbACRcqRrMQaziNQCIRMbCGkxl8h99uDjDuvEBrHDzZ8L99il9087/PwBQ41NNC4tljPSzQ1nnklPbJBObJOe8Sw5d8btNMMGhmNpubisMriRxibN23YYrIWZ5wBzbsPtBMMGxobD+LX8hWhp/sxzoD5d/nb6R8JaGCFaGqR5DhxOMLidg98Wy55nbJK9benGYL/wHDhuuPF2TuGMBDx+MWdPY5P42WYt2yCd/hjosGp5udvpGz58qLHB7TAYw/4AsnACDuUoWkbBKBgFo2AU4AQAcRZbZULvRMMAAAAASUVORK5CYII=","orcid":"","institution":"Applied Materials","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Purnendu","middleName":"","lastName":"Mishra","suffix":""},{"id":269926830,"identity":"04c098f1-3e92-46d6-b0cc-3b878123d22c","order_by":1,"name":"Akash Modi","email":"","orcid":"","institution":"Applied Materials","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Akash","middleName":"","lastName":"Modi","suffix":""},{"id":269926831,"identity":"94289817-bd85-4a37-8800-2b99c308c246","order_by":2,"name":"Sumit Jha","email":"","orcid":"","institution":"Applied Materials","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Sumit","middleName":"","lastName":"Jha","suffix":""},{"id":269926832,"identity":"772b1684-159c-4193-82cd-5bc8d1ec8656","order_by":3,"name":"Rajiv Kumar","email":"","orcid":"","institution":"Applied Materials","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Rajiv","middleName":"","lastName":"Kumar","suffix":""},{"id":269926833,"identity":"8dd32335-94ba-4ec0-b0bf-8b19b5dbd572","order_by":4,"name":"Kiran Aatre","email":"","orcid":"","institution":"Applied Materials","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Kiran","middleName":"","lastName":"Aatre","suffix":""},{"id":269926834,"identity":"00419b3a-fa71-45c8-a8c4-b969d18023e0","order_by":5,"name":"Gursewak Singh","email":"","orcid":"","institution":"Applied Materials","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Gursewak","middleName":"","lastName":"Singh","suffix":""},{"id":269926835,"identity":"0b790843-8a1d-433d-a149-fa1216330013","order_by":6,"name":"Shubham Mathur","email":"","orcid":"","institution":"Applied Materials","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Shubham","middleName":"","lastName":"Mathur","suffix":""}],"badges":[],"createdAt":"2024-01-29 11:06:13","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3908756/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3908756/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":50774031,"identity":"1181f4ac-521e-4e87-9352-8282260b2f26","added_by":"auto","created_at":"2024-02-07 05:40:45","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1754318,"visible":true,"origin":"","legend":"","description":"","filename":"roisubmittedv1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3908756/v1_covered_d1c12227-840a-4ee6-9700-802af6d12425.pdf"}],"financialInterests":"(Not answered)","formattedTitle":"Multi-Stain Multi-Level Convolutional Network for Multi-Tissue breast cancer image segmentation","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","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":"","lastPublishedDoi":"10.21203/rs.3.rs-3908756/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3908756/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Digital pathology and microscopy image analysis are widely employed in the segmentation of digitally scanned IHC slides, primarily to identify cancer and pinpoint regions of interest (ROI) indicative of tumor presence. However, current ROI segmentation models are either stain specific or suffer from the issues of stain and scanner variance due to different staining protocols or modalities across multiple labs. Also, tissues like Ductal Carcinoma in Situ (DCIS), ACINIs etc. are often classified as Tumors due to their structural similarities and color compositions. In this paper, we proposed a novel convolutional neural network (CNN) based Multi-class Tissue Segmentation model for histopathology whole-slide Breast slides which classifies tumor and segments other tissue regions such as Ducts, ACINIs, DCIS, Squamous epithelium, Blood Vessels, Necrosis etc. as separate class. Our unique pixel aligned non-linear merge across spatial resolutions, empowers models with both local and global field of view for accurate detection of various classes. Our proposed model is also able separate bad regions such as folds, artifacts, blurry regions, bubbles etc. from tissue regions using multi-level context from different resolutions of WSI. Multi-phase iterative training with context aware augmentation and increasing noise was used to efficiently train multi-stain generic model with partial and noisy annotations from 513 slides. Our training pipeline used 12 million patches generated using context aware augmentations which made our model stain \u0026 scanner invariant across data sources. We have achieved an Area Under curve (AUC) of 0.85 for tumor and 0.93 for other class on 35000 test patches distributed across stains, scanners and data sources. To extrapolate stain \u0026 scanner invariance, our model was evaluated on 23000 patches which were for a completely new stain (Hematoxylin and Eosin) from a completely new scanner (Motic) from a different lab. The mean IOU was 0.72 which is on par with model performance on other data sources and scanners. Our results illustrate that the proposed model exhibits remarkable accuracy with consistent performance and broad utility especially in facilitating pathologists’ assessment of breast cancer.","manuscriptTitle":"Multi-Stain Multi-Level Convolutional Network for Multi-Tissue breast cancer image segmentation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-02-06 10:24:13","doi":"10.21203/rs.3.rs-3908756/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":"5c86c174-a959-449b-8bcd-b154b3fb9125","owner":[],"postedDate":"February 6th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":28438309,"name":"Biological sciences/Computational biology and bioinformatics/Machine learning"},{"id":28438310,"name":"Biological sciences/Computational biology and bioinformatics/Image processing"},{"id":28438311,"name":"Biological sciences/Cancer/Breast cancer"}],"tags":[],"updatedAt":"2024-02-19T18:06:41+00:00","versionOfRecord":[],"versionCreatedAt":"2024-02-06 10:24:13","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-3908756","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3908756","identity":"rs-3908756","version":["v1"]},"buildId":"cTy_lsJlmDsVRNrSptgXS","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}