[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-339759-105":59,"doc-detail-339759-en":131},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":124,"head_meta":126,"extra_data":128,"updated_unix":130},105,"en","robust-credible-and-interpretable-ai-based-histopathological-prostate-cancer-grading","Robust, credible, and interpretable AI-based histopathological prostate cancer grading","","Prostate cancer diagnosis depends on histopathological biopsy assessment, where AI grading models can degrade on external data due to differences in sample preparation and to ISUP-grade prediction that may propagate human annotation errors. The work introduces the prostate cancer aggressiveness index (PCAI), a detection and grading framework trained on objective patient outcomes. PCAI is built to provide robustness to data variation, medical interpretability, and prediction confidence, and is evaluated across multicentric retrospective cohorts with external testing. Results quantify performance drops from slide and staining variability and show improved, above-ISUP performance after targeted algorithmic upgrades, including domain adversarial training.",{"@graph":69,"@context":123},[70,84,106],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/robust-credible-and-interpretable-ai-based-histopathological-prostate-cancer-grading/339759/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/robust-credible-and-interpretable-ai-based-histopathological-prostate-cancer-grading/339759.png","ImageObject",300,407,{"name":92,"@type":93},"Liam","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-11-12","2026-09-21",true,{"@type":102,"interactionType":103,"userInteractionCount":105},"InteractionCounter",{"@type":104},"ViewAction",13,{"@type":107,"mainEntity":108},"FAQPage",[109,115,119],{"name":110,"@type":111,"acceptedAnswer":112},"Why do AI-based prostate cancer grading models often perform worse on external datasets?","Question",{"text":113,"@type":114},"Reduced performance can be caused by variations in sample preparation such as staining protocol, section thickness, and the scanner used. Another issue is ISUP-grade prediction, which can perpetuate human annotation errors.","Answer",{"name":116,"@type":111,"acceptedAnswer":117},"What is the prostate cancer aggressiveness index (PCAI)?",{"text":118,"@type":114},"PCAI is an AI-based prostate cancer detection and grading framework trained on objective patient outcome instead of subjective ISUP grades. It includes modules aimed at robustness to data variation, medical interpretability, and prediction confidence.",{"name":120,"@type":111,"acceptedAnswer":121},"What improvements does the study report after addressing data variation and model credibility?",{"text":122,"@type":114},"Differences in slide processing can significantly reduce AI grading performance. Targeted algorithmic improvements, including domain adversarial training, improve robustness and credibility, leading to prediction performance that systematically exceeds expert ISUP grading across multiple biopsy and TMA cohorts.","https://schema.org",{"og:url":83,"og:type":125,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":127,"canonical":83},"index,follow",{"doc_id":129,"site_id":62},339759,1790479958,{"code":4,"msg":5,"data":132},{"doc_id":129,"user_id":133,"nickname":92,"user_avatar":134,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":135,"file_id":136,"file_url":137,"file_type":138,"file_size":139,"view_count":105,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":140,"language":141,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":142,"faqs":143,"seo_title":144,"seo_description":67,"update_tm":145,"read_time":146,"preview_page_count":19,"preview_urls":147},8796095461564,"https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d","medRxiv preprint doi: [https://doi.org/10.1101/2024.07.09.24310082](https://doi.org/10.1101/2024.07.09.24310082); this version posted July 10, 2024. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  \nIt is made available under a CC-BY-NC-ND 4.0 International license .  \nRobust, credible, and interpretable AI-based histopathological prostate  \ncancer grading  \nFabian Westhaeusser 1,2†, Patrick Fuhlert 1,2†, Esther Dietrich&1, Maximilian Lennartz&3, Robin Khatri&1, Nico Kaiser&1,4, Pontus Röbeck&5, Roman Bülow6, Saskia von Stillfried6, Anja Witte 1, Sam Ladjevardi5, Anders Drotte2, Peter Severgardh2, Jan Baumbach7, Victor G. Puelles4,8,9, Michael Häggman5, Michael Brehler 1, Peter Boor6, Peter Walhagen2, Anca Dragomir 10, Christer Busch2,5, Markus Graefen 11, Ewert Bengtsson2,12, Guido Sauter%3, Marina Zimmermann%1, Stefan Bonn%1,2*  \n1Institute of Medical Systems Biology, Center for Biomedical AI (bAIome), Center for Molecular Neurobiology Hamburg (ZMNH), University Medical Center Hamburg-Eppendorf, Hamburg, Germany.  \n2 Spearpoint Analytics AB, Stockholm, Sweden.  \n3Institute of Pathology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany.  \n4III. Department of Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany.  \n5Department of Urology, Uppsala University Hospital, Uppsala, Sweden.  \n6Institute of Pathology, RWTH Aachen University Hospital, Aachen, Germany.  \n7Institute of Computational Systems Biology, University of Hamburg, Germany.  \n8Department of Clinical Medicine, Aarhus University, Aarhus, Denmark.  \n9Department of Pathology, Aarhus University Hospital, Aarhus, Denmark.  \n10Department of Pathology, Uppsala University Hospital and Department of Immunology, Genetics11Martini-and PatKlinikhPorogystat,eUppsCanaca Unr Ceverter,sityUn,iUpverpssitalyaH, SowedespitalnH. amburg-Eppendorf, Hamburg, G12ermanyUppsala. University, Department of Information Technology, Centre for Image Analysis, Uppsala, Sweden.  \n*Corresponding author: [sbonn@uke.de](sbonn@uke.de) ;  \n†,&,%These authors contributed equally to this work.  \nNOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.  \nmedRxiv preprint doi: [https://doi.org/10.1101/2024.07.09.24310082](https://doi.org/10.1101/2024.07.09.24310082); this version posted July 10, 2024. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  \nIt is made available under a CC-BY-NC-ND 4.0 International license .  \nSummary  \nBackground: Prostate cancer (PCa) is among the most common cancers in men and its diagnosis requires the histopathological evaluation of biopsies by human experts. While several recent artificial intelligence-based (AI) approaches have reached human expert-level PCa grading, they often display significantly reduced performance on external datasets. This reduced performance can be caused by variations in sample preparation, for instance the staining protocol, section thickness, or scanner used. Another limiting factor of contemporary AI-based PCa grading is the prediction of ISUP grades, which leads to the perpetuation of human annotation errors.  \nMethods: We developed the prostate cancer aggressiveness index (PCAI), an AI-based PCa detection and grading framework that is trained on objective patient outcome, rather than subjective ISUP grades. We designed PCAI as a clinical application, containing algorithmic modules that offer robustness to data variation, medical interpretability, and a measure of prediction confidence. To train and evaluate PCAI, we generated a multicentric, retrospective, observational trial consisting of six cohorts with 25,591 patients, 83,864 images, and 5 years of median follow-up from 5 different c","cbCailPY23UMZOKA","https://ap.wps.com/l/cbCailPY23UMZOKA","pdf",4174785,25,"English","# Summary\n## Background\n## Methods\n## Findings\n## Interpretation\n# Keywords","[{\"question\":\"Why do AI-based prostate cancer grading models often perform worse on external datasets?\",\"answer\":\"Reduced performance can be caused by variations in sample preparation such as staining protocol, section thickness, and the scanner used. Another issue is ISUP-grade prediction, which can perpetuate human annotation errors.\"},{\"question\":\"What is the prostate cancer aggressiveness index (PCAI)?\",\"answer\":\"PCAI is an AI-based prostate cancer detection and grading framework trained on objective patient outcome instead of subjective ISUP grades. It includes modules aimed at robustness to data variation, medical interpretability, and prediction confidence.\"},{\"question\":\"What improvements does the study report after addressing data variation and model credibility?\",\"answer\":\"Differences in slide processing can significantly reduce AI grading performance. Targeted algorithmic improvements, including domain adversarial training, improve robustness and credibility, leading to prediction performance that systematically exceeds expert ISUP grading across multiple biopsy and TMA cohorts.\"}]","Robust, credible, and interpretable AI-based histopathological prostate cancer grading | PDF",1790032053,63,[148,152,156,160],{"page_no":8,"url":149,"width":150,"height":151},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/7b13cbd35c03d585e46c56657b456124.webp",983,1388,{"page_no":14,"url":153,"width":154,"height":155},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/c27d4979769ecfd5e9fc5dd76da6b9d9.webp",982,1387,{"page_no":81,"url":157,"width":158,"height":159},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/8a38fe23ad6765a4fdee4701a475dbc6.webp",981,1385,{"page_no":19,"url":161,"width":162,"height":163},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/f6e7d559c628cc0e01ce6fb0a359b3b2.webp",980,1384]