[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-367617-105":59,"doc-detail-367617-en":130},{"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":123,"head_meta":125,"extra_data":127,"updated_unix":129},105,"en","generative-ai-for-weakly-supervised-segmentation-and-downstream-classification-of-brain-tumors-on-mr-images","Generative AI for weakly supervised segmentation and downstream classification of brain tumors on MR images","","Generative AI is applied to medical image analysis to address the limited availability of manual tumor annotations for supervised brain-tumor segmentation on MR images. The approach uses simpler binary image-level labels to generate healthy variants of cancerous images as priors for training. Because weak supervision can yield occasional unreliable segmentations, generated non-cancerous variants are used to select the most effective segmentations without ground truth. Performance reaches Dice 79.27% (BraTS 2020) and 73.58% (pLGG) and improves further after removing suboptimal segmentations, with tumor classification AUCs reaching 93.54% and 83.74%.",{"@graph":69,"@context":122},[70,84,105],{"@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":35,"@type":76,"position":81},"https://docshare.wps.com/document/healthcare/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/generative-ai-for-weakly-supervised-segmentation-and-downstream-classification-of-brain-tumors-on-mr-images/367617/",{"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/generative-ai-for-weakly-supervised-segmentation-and-downstream-classification-of-brain-tumors-on-mr-images/367617.png","ImageObject",300,407,{"name":92,"@type":93},"Chumphorn","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-24","2026-09-23",true,{"@type":102,"interactionType":103,"userInteractionCount":24},"InteractionCounter",{"@type":104},"ViewAction",{"@type":106,"mainEntity":107},"FAQPage",[108,114,118],{"name":109,"@type":110,"acceptedAnswer":111},"Why are manual annotations a limitation for brain tumor segmentation on MR images?","Question",{"text":112,"@type":113},"Manual segmentation requires extensive time and resources from radiologists and is often difficult to obtain, especially for specific contexts such as pediatric cancer.","Answer",{"name":115,"@type":110,"acceptedAnswer":116},"How does the proposed weakly supervised method train without ground-truth masks?",{"text":117,"@type":113},"It uses simpler binary image-level labels and generates healthy variants of cancerous images to serve as priors for training the segmentation model.",{"name":119,"@type":110,"acceptedAnswer":120},"How are unreliable weakly supervised segmentations handled for downstream classification?",{"text":121,"@type":113},"The method uses generated non-cancerous variants to identify and remove suboptimal segmentations without needing ground-truth annotations.","https://schema.org",{"og:url":83,"og:type":124,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":126,"canonical":83},"index,follow",{"doc_id":128,"site_id":62},367617,1790301842,{"code":4,"msg":5,"data":131},{"doc_id":128,"user_id":132,"nickname":92,"user_avatar":133,"doc_module":4,"category_id":34,"category_name":35,"doc_title":65,"doc_description":67,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":24,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":139,"language":140,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":67,"update_tm":144,"read_time":145},2336475401981,"https://ap-avatar.wpscdn.com/avatar/22000c94efd8d5204d?x-image-process=image/resize,m_fixed,w_180,h_180&k=1786935347598174694","[www. nature.com/scientificreports](www. nature.com/scientificreports)  \nOPEN  \nGenerative AI for weakly supervised segmentation and downstream classification of brain tumors on MR images  \nJay J. Yoo1,2,4,6, Khashayar Namdar1,2,6, Matthias W. Wagner2,3,12, Kristen W. Yeom13,14, Liana F. Nobre15, Uri Tabori7,8,9, Cynthia Hawkins8,10,11, Birgit B. Ertl-Wagner2,3 & Farzad Khalvati1,2,3,4,5,6􀀍  \nSegmenting abnormalities is a leading problem in medical imaging. Using machine learning for segmentation generally requires manually annotated segmentations, demanding extensive time and resources from radiologists. We propose a weakly supervised approach that utilizes binary image-level labels, which are much simpler to acquire, rather than manual annotations to segment brain tumors on magnetic resonance images. The proposed method generates healthy variants of cancerous images for use as priors when training the segmentation model. However, using weakly supervised segmentations for downstream tasks such as classification can be challenging due to occasional unreliable segmentations. To address this, we propose using the generated non-cancerous variants to identify the most effective segmentations without requiring ground truths. Our proposed method generates segmentations that achieve Dice coefficients of 79.27% on the Multimodal Brain Tumor Segmentation (BraTS) 2020 dataset and 73.58% on an internal dataset of pediatric low-grade glioma (pLGG), which increase to 88.69% and 80.29%, respectively, when removing suboptimal segmentations identified using the proposed method. Using the segmentations for tumor classification results with Area Under the Characteristic Operating Curve (AUC) of 93.54% and 83.74% on the BraTS and pLGG datasets, respectively. These are comparable to using manual annotations which achieve AUCs of 95.80% and 83.03% on the BraTS and pLGG datasets, respectively.  \nSegmenting anomalies in medical images is a leading problem as it can be used to assist in automated diagnosis1. However, manually segmenting anomalies demands extensive time and resources from radiologists. Training Machine Learning (ML) models for this segmentation task in a supervised manner requires ground-truth manual segmentations that are also cumbersome to acquire2,3. Furthermore, segmentation models often demand large datasets of annotated medical images, which may not be available for specific contexts such as pediatric cancer. Hence, training ML models with limited or without ground truth annotations masks is key.  \nGhosal et al. proposed a compound attention-embedded dual-channel encoder-decoder model for multiple sclerosis (MS) lesion segmentation in brain MRI4. This approach utilizes residual connections and channel-wise attention to effectively capture both local and global features, achieving a Dice coefficient of 0.73 on the MICCAI  \n1University of Toronto, Institute of Medical Science, Toronto M5S 1A8, Canada. 2Department of Diagnostic & Interventional Radiology, The Hospital for Sick Children, Toronto M5G 1X8, Canada. 3Department of Medical Imaging, University of Toronto, Toronto M5S 1A8, Canada. 4Department of Computer Science, University of Toronto, Toronto M5S 1A8, Canada. 5Department of Mechanical and Industrial Engineering, University of Toronto, Toronto M5S 1A8, Canada. 6Vector Institute, Toronto M5G 1M1, Canada. 7Division of Hematology/Oncology, The Hospital for Sick Children, Toronto M5G 1X8, Canada. 8Developmental and Stem Cell Biology Program, The Hospital for Sick Children, Toronto M5G 1X8, Canada. 9Department of Medical Biophysics, University of Toronto, Toronto M5S 1A8, Canada. 10Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto M5S 1A8, Canada. 11Department of Paediatric Laboratory Medicine, The Hospital for Sick Children, Toronto M5G 1X8, Canada. 12University Hospital Augsburg, Institute of Diagnostic and Interventional Neuroradiology, Augsburg 86156, Germany. 13Department of Radiology, Stan","cbCaifz2MRijGOcV","https://ap.wps.com/l/cbCaifz2MRijGOcV","pdf",4221613,21,"English","# Generative AI for weakly supervised segmentation and downstream classification of brain tumors on MR images\n## Problem motivation and limitations of fully supervised segmentation\n## Prior work on segmentation in brain MRI and other medical imaging\n## Proposed weakly supervised framework using binary labels and generated variants\n## Segmentation and classification performance on BraTS and pediatric pLGG datasets","[{\"question\":\"Why are manual annotations a limitation for brain tumor segmentation on MR images?\",\"answer\":\"Manual segmentation requires extensive time and resources from radiologists and is often difficult to obtain, especially for specific contexts such as pediatric cancer.\"},{\"question\":\"How does the proposed weakly supervised method train without ground-truth masks?\",\"answer\":\"It uses simpler binary image-level labels and generates healthy variants of cancerous images to serve as priors for training the segmentation model.\"},{\"question\":\"How are unreliable weakly supervised segmentations handled for downstream classification?\",\"answer\":\"The method uses generated non-cancerous variants to identify and remove suboptimal segmentations without needing ground-truth annotations.\"}]","Generative AI for weakly supervised segmentation and downstream classification of brain tumors on MR images | PDF",1790171644,53]