[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-83119-en":3,"doc-seo-83119-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},83119,1099514067438,"River Wang","https://ap-avatar.wpscdn.com/avatar/100002539ee87300030?x-image-process=image/resize,m_fixed,w_180,h_180&k=1780474512215547542",8,"Research & Report","Unsupervised Domain Adaptation for Calcification Classification in Mammography Across Multi-Site Datasets","Deep learning-based computer-aided diagnosis systems can classify breast disease on mammography, yet performance degrades when data come from unseen sites due to domain shifts. The study proposes an unsupervised calcification classification framework for malignant versus benign disease across multi-site datasets. It combines a style-transfer domain adaptation module using AdaIN and CycleGAN to create vendor/technique-specific training samples without extra annotations, with a Swin Transformer V2 supervised classifier and multi-dataset evaluation with AUC gains on EMBED and Duke data.","GENERIC COLORIZED JOURNAL, VOL. XX, NO. XX, XXXX 2023 1  \nUnsupervised Domain Adaptation for Calcification Classification in Mammography Across Multi-Site Datasets  \nXuan Liu, Derek L. Nguyen, Emily C. Barre, Jennifer Thomas, Thomas Lynch, Jeffrey R. Marks, E. Shelley  \nHwang, Marc D. Ryser, Joseph Y. Lo, Lars J. Grimm  \narXiv :2607 .06549v 1 [ cs .CV] 7 Jul 2026  \nAbstract—Deep learning-based computer-aided diagnosis (CAD) systems have shown strong performance in breast cancer diagnosis, particularly for classification tasks in mammography. However, domain shifts across multi-site datasets remain a challenge, especially when models are applied to unseen domains. In this work, we proposed a calcification classification framework to improve malignant versus benign breast disease classification across multi-site mammography datasets. The framework consisted of two components: (1) an unsupervised domain adaptation module based on style transfer models (AdaIN and CycleGAN) to generate vendor-specific and technique-specific training samples without additional annotations, and (2) a supervised classification module using Swin Transformer V2 as the backbone. We evaluated the proposed method on three datasets: crossvalidation on OPTIMAM (National Health Service, United Kingdom; n=2994), followed by external validation on EMBED (Emory University; n=125), and Duke Calcification Dataset v1 (n=788). These datasets cover multiple vendors and include both full-field digital mammography and synthetic 2D images derived from digital breast tomosynthesis. The proposed framework improved cross-site performance for both EMBED (AUC 0.68 to 0.72) and the Duke Calcification Dataset (AUC 0.68 to 0.73). These findings indi  \nThis study was supported in part by NIH/NCI R01 CA271237 . The mammography images and data used in this research were derived from the OPTIMAM imaging database (OMI-DB)([https://medphys.royalsurrey.nhs.uk/omidb](https://medphys.royalsurrey.nhs.uk/omidb)). We would like to acknowledge the OPTIMAM project team and staff at the Royal Surrey NHS Foundation Trust who developed the OPTIMAM database, Cancer Research UK which funded the creation and maintenance of OPTIMAM database and Cancer Research Horizons which facilitates access to the OPTIMAM data. (Corresponding author: Lars J. Grimm)  \nXuan Liu is with the Department of Electrical and Computer Engineering and the Department of Radiology, Duke University, Durham, NC, USA ([e-mail: xuan.liu115@duke.edu](e-mail: xuan.liu115@duke.edu)) .  \nDerek L. Nguyen, Joseph Y. Lo and Lars J. Grimm are with the Department of Radiology, Duke University, Durham, NC, USA (e-mail: [derek.nguyen@duke.edu](derek.nguyen@duke.edu) ; [joseph.lo@duke.edu](joseph.lo@duke.edu) ; [lars.grimm@duke.edu](lars.grimm@duke.edu)).  \nEmily C. Barre is with the Duke University School of Medicine, Durham, NC, USA (e-mail: [emily.barre@duke.edu](emily.barre@duke.edu)).  \nJennifer Thomas is with the Department of Population Health Sciences, Duke University, Durham, NC, USA (e-mail: [jennifer.grant@duke.edu](jennifer.grant@duke.edu)).  \nThomas Lynch, Jeffrey R. Marks and E. Shelley Hwang are with the Department of Surgery, Duke University, Durham, NC, USA (e-mail: [thomas.lynch2@duke.edu](thomas.lynch2@duke.edu) ; [jeffrey.marks@duke.edu](jeffrey.marks@duke.edu) ; shel  \n[ley.hwang@duke.edu](ley.hwang@duke.edu)).  \nMarc D. Ryser is with the Department of Population Health Sciences, Duke University, Durham, NC, USA and the Faculty of Medicine, University of Geneva, Geneva, Switzerland ([e-mail: marc.ryser@duke.edu](e-mail: marc.ryser@duke.edu)).  \ncate that domain adaptation can reduce domain shifts and improve the generalization for calcification classification across multi-site datasets.  \nIndex Terms—Classification, Domain Adaptation, Style Transfer, Mammography, Computer-aided Diagnosis  \nI. INTRODUCTION  \nBREAST cancer is one of the leading causes of cancer  \nrelated deaths among women in the United States, highlighti","cbCaihllPXOlsUhc","https://ap.wps.com/l/cbCaihllPXOlsUhc","pdf",15052106,3,1,10,"English","en",105,"# Abstract\n# Index Terms\n# I. Introduction","[{\"question\":\"What problem does the proposed framework address in mammography-based calcification classification?\",\"answer\":\"It targets cross-site domain shifts that reduce model generalization when mammography data come from unseen vendors, techniques, or acquisition settings.\"},{\"question\":\"How does the method perform unsupervised domain adaptation?\",\"answer\":\"It uses style transfer models—AdaIN and CycleGAN—to generate vendor-specific and technique-specific training samples without requiring additional annotations.\"},{\"question\":\"Which model is used for the supervised classification component and how was it evaluated?\",\"answer\":\"A Swin Transformer V2 backbone performs supervised malignant-versus-benign classification, evaluated via cross-validation on OPTIMAM and external validation on EMBED and the Duke Calcification Dataset v1.\"}]",1784185409,25,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"unsupervised-domain-adaptation-for-calcification-classification-in-mammography-across-multi-site-datasets","",{"@graph":36,"@context":85},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,50],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":20},"https://docshare.wps.com/document/research-report/",{"item":51,"name":13,"@type":43,"position":52},"https://docshare.wps.com/document/unsupervised-domain-adaptation-for-calcification-classification-in-mammography-across-multi-site-datasets/83119/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":24,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-25","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does the proposed framework address in mammography-based calcification classification?","Question",{"text":75,"@type":76},"It targets cross-site domain shifts that reduce model generalization when mammography data come from unseen vendors, techniques, or acquisition settings.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the method perform unsupervised domain adaptation?",{"text":80,"@type":76},"It uses style transfer models—AdaIN and CycleGAN—to generate vendor-specific and technique-specific training samples without requiring additional annotations.",{"name":82,"@type":73,"acceptedAnswer":83},"Which model is used for the supervised classification component and how was it evaluated?",{"text":84,"@type":76},"A Swin Transformer V2 backbone performs supervised malignant-versus-benign classification, evaluated via cross-validation on OPTIMAM and external validation on EMBED and the Duke Calcification Dataset v1.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,134],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":22,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":22,"slug":133},"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]