[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-332941-105":3,"detail-sidebar-cat-0-en-105":81,"doc-detail-332941-en":130},{"code":4,"msg":5,"data":6},0,"ok",{"site_id":7,"language":8,"slug":9,"title":10,"keywords":11,"description":12,"schema_data":13,"social_meta":74,"head_meta":76,"extra_data":78,"updated_unix":80},105,"en","stress-testing-deep-learning-models-for-prostate-cancer-detection-on-biopsies-and-surgical-specimens","Stress testing deep learning models for prostate cancer detection on biopsies and surgical specimens","","Deep learning aims to support pathologists in assessing the presence, location, and extent of prostate cancer on H&E-stained whole-slide images from biopsies and radical prostatectomies. This study evaluates whether morphological differences between specimen types degrade performance when biopsy-trained CNNs are applied to prostatectomy slides, and vice versa. Using DenseNet models across multiple cohorts (N=1,000) and external digitized biopsy data (N=750), performance drops across sample types and morphologically driven decision making limits generalizability. Results indicate the need for tailored, sample-specific machine learning models.",{"@graph":14,"@context":73},[15,34,56],{"@type":16,"itemListElement":17},"BreadcrumbList",[18,23,27,31],{"item":19,"name":20,"@type":21,"position":22},"https://docshare.wps.com","Home","ListItem",1,{"item":24,"name":25,"@type":21,"position":26},"https://docshare.wps.com/document/","Document",2,{"item":28,"name":29,"@type":21,"position":30},"https://docshare.wps.com/document/research-report/","Research & Report",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/stress-testing-deep-learning-models-for-prostate-cancer-detection-on-biopsies-and-surgical-specimens/332941/",4,{"url":32,"name":10,"@type":35,"image":36,"author":41,"headline":10,"publisher":44,"fileFormat":47,"inLanguage":8,"description":12,"dateModified":48,"datePublished":49,"encodingFormat":47,"isAccessibleForFree":50,"interactionStatistic":51},"DigitalDocument",{"url":37,"@type":38,"width":39,"height":40},"https://docshare.wps.com/thumbnails/stress-testing-deep-learning-models-for-prostate-cancer-detection-on-biopsies-and-surgical-specimens/332941.png","ImageObject",300,407,{"name":42,"@type":43},"Theodora","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-31","2026-09-21",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",8,{"@type":57,"mainEntity":58},"FAQPage",[59,65,69],{"name":60,"@type":61,"acceptedAnswer":62},"What is the main objective of the study?","Question",{"text":63,"@type":64},"To determine whether morphological differences between prostate biopsy and radical prostatectomy specimens affect the performance of biopsy-trained deep learning cancer detection models when applied across sample types.","Answer",{"name":66,"@type":61,"acceptedAnswer":67},"How were the deep learning models trained and evaluated?",{"text":68,"@type":64},"A DenseNet CNN was trained and validated using biopsy and radical prostatectomy datasets, with an additional combined dataset, using multiple cohorts totaling N=1,000 and a tile-level and whole-slide evaluation.",{"name":70,"@type":61,"acceptedAnswer":71},"What did the external validation show?",{"text":72,"@type":64},"External validation using digitized biopsy slide images from an RTOG clinical trial (N=750) confirmed performance differences between models across specimen types, supported by qualitative and quantitative analyses.","https://schema.org",{"og:url":32,"og:type":75,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":77,"canonical":32},"index,follow",{"doc_id":79,"site_id":7},332941,1790216294,{"code":4,"msg":82,"data":83},"success",[84,88,92,96,101,106,111,114,119,122,126],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":85,"show_sort_weight":86,"slug":87},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":89,"show_sort_weight":90,"slug":91},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":93,"show_sort_weight":94,"slug":95},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":25,"category_name":98,"show_sort_weight":99,"slug":100},5,"Comic",60,"comic",{"id":102,"doc_module":4,"doc_module_name":25,"category_name":103,"show_sort_weight":104,"slug":105},6,"Technology",50,"technology",{"id":107,"doc_module":4,"doc_module_name":25,"category_name":108,"show_sort_weight":109,"slug":110},7,"Healthcare",40,"healthcare",{"id":55,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":112,"slug":113},30,"research-report",{"id":115,"doc_module":4,"doc_module_name":25,"category_name":116,"show_sort_weight":117,"slug":118},9,"Religion & Spirituality",20,"religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":25,"category_name":120,"show_sort_weight":117,"slug":121},"World Cup","world-cup",{"id":123,"doc_module":4,"doc_module_name":25,"category_name":124,"show_sort_weight":123,"slug":125},10,"Lifestyle","lifestyle",{"id":127,"doc_module":4,"doc_module_name":25,"category_name":128,"show_sort_weight":97,"slug":129},19,"General","general",{"code":4,"msg":82,"data":131},{"doc_id":79,"user_id":132,"nickname":42,"user_avatar":133,"doc_module":4,"category_id":55,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":139,"language":140,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":12,"update_tm":144,"read_time":112,"preview_page_count":4},687197207919,"https://ap-avatar.wpscdn.com/avatar/a000253d6f5f7c60be?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779446848396160552","Journal of Pathology  \nJ Pathol February 2025; 265: 146–157  \nPublished online 11 December 2024 in Wiley Online Library (wi[leyonlinelibrary.com](leyonlinelibrary.com) DOI: 10.1002/path.6373)[)](leyonlinelibrary.com) DOI: 10.1002/path.6373)[ DOI:](leyonlinelibrary.com) DOI: 10.1002/path.6373)[ 10.1002/path.6373](leyonlinelibrary.com) DOI: 10.1002/path.6373)  \nORIGINAL ARTICLE  \nStress testing deep learning models for prostate cancer detection on biopsies and surgical specimens  \nBrennan Tuomas  \nT Flannery 1 , Howard M Sandler2, Priti Lal3, Mirtti6, Xavier Farre7, Rohann Correa8, Susan  \nMichael D Feldman4, Juan C Santa-Rosario5, Tilak Pathak6, Chafe9, Amit Shah 10, Jason A Efstathiou 11 , Karen Hoffman 12,  \nMark A Hallman 13, Michael Straza 14, Richard Jordan 15, Stephanie L Pugh 16, Felix Feng 17 and Anant Madabhushi6,18*  \n1 Case Western Reserve University, Cleveland, OH, USA  \n2 Cedars-Sinai Medical Center, Los Angeles, CA, USA  \n3 University of Pennsylvania, Philadelphia, PA, USA  \n4 Indiana University, Indianapolis, IN, USA  \n5 CorePlus, Carolina, Puerto Rico  \n6 Emory Winship Cancer Institute, Atlanta, GA, USA  \n7 Public Health Agency of Catalonia, Catalonia, Spain  \n8 London Health Science Centre, London, ON, Canada  \n9 Cross Cancer Institute, Edmonton, AB, Canada  \n10 WellSpan Health, York, PA, USA  \n11 Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA  \n12 The University of Texas MD Anderson Cancer Center, Houston, TX, USA  \n13 Fox Chase Cancer Center, Philadelphia, PA, USA  \n14 Medical College of Wisconsin, Milwaukee, WI, USA  \n15 NRG Oncology Biospecimen Bank, San Francisco, CA, USA  \n16 NRG Oncology Statistics and Data Management Center, Philadelphia, PA, USA  \n17 University of California San Francisco, San Francisco, CA, USA  \n18 Atlanta Veterans Affairs Medical Center, Atlanta, GA, USA  \n*Correspondence to: A Madabhushi, Atlanta Veterans Affairs Medical Center, Atlanta, GA, USA. E-mail: [anantm@emory.edu](anantm@emory.edu)  \nAbstract  \nThe presence, location, and extent of prostate cancer is assessed by pathologists using H&E-stained tissue slides.  \nMachine learning approaches can accomplish these tasks for both biopsies and radical prostatectomies. Deep learning approaches using convolutional neural networks (CNNs) have been shown to identify cancer in pathologic slides, some securing regulatory approval for clinical use. However, differences in sample processing can subtly alter the morphology between sample types, making it unclear whether deep learning algorithms will consistently work on both types of slide images. Our goal was to investigate whether morphological differences between sample types affected the performance of biopsy-trained cancer detection CNN models when applied to radical prostatectomiesand vice versa using multiple cohorts (N = 1,000). Radical prostatectomies (N = 100) and biopsies (N = 50) were acquired from The University of Pennsylvania to train (80%) and validate (20%) a DenseNet CNN for biopsies (MB), radical prostatectomies (MR), and a combined dataset (MB+R). On a tile level, MB and MR achieved F1 scores greater than 0.88 when applied to their own sample type but less than 0.65 when applied across sample types. On a wholeslide level, models achieved significantly better performance on their own sample type compared to the alternative model (p \u003C 0.05) for all metrics. This was confirmed by external validation using digitized biopsy slide images from a clinical trial [NRG Radiation Therapy Oncology Group (RTOG)](NRG/RTOG 0521, N = 750) via both qualitative and quantitative analyses (p \u003C 0.05). A comprehensive review of model outputs revealed morphologically driven decision making that adversely affected model performance. MB appeared to be challenged with the analysis of open gland structures, whereas MR appeared to be challenged with closed gland structures, indicating potential morphological variation between the training sets. These findings suggest that diffe","cbCaic71CeGu1Qmo","https://ap.wps.com/l/cbCaic71CeGu1Qmo","pdf",8638726,12,"English","# Abstract\n## Study aim\n## Data and model training\n## Performance results\n## External validation\n## Error analysis and implications","[{\"question\":\"What is the main objective of the study?\",\"answer\":\"To determine whether morphological differences between prostate biopsy and radical prostatectomy specimens affect the performance of biopsy-trained deep learning cancer detection models when applied across sample types.\"},{\"question\":\"How were the deep learning models trained and evaluated?\",\"answer\":\"A DenseNet CNN was trained and validated using biopsy and radical prostatectomy datasets, with an additional combined dataset, using multiple cohorts totaling N=1,000 and a tile-level and whole-slide evaluation.\"},{\"question\":\"What did the external validation show?\",\"answer\":\"External validation using digitized biopsy slide images from an RTOG clinical trial (N=750) confirmed performance differences between models across specimen types, supported by qualitative and quantitative analyses.\"}]","Stress testing deep learning models for prostate cancer detection on biopsies and surgical specimens | PDF",1790000076]