[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82935-en":3,"doc-seo-82935-105":29,"detail-sidebar-cat-0-en-105":90},{"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":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},82935,8796095461610,"Oliver","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","CenSynCMB Centre Maps and Physics-Guided Synthesis for Microbleed Detection","Cerebral microbleeds (CMBs) serve as MRI markers of small vessel disease and the microbleed component of ARIA-H, yet automated detection remains difficult due to their small size, sparse distribution, and visual similarity to vessels, calcification-like foci, and artefacts. CenSynCMB introduces a centre-guided, mimic-aware framework combining a 3D Attention U-Net, centre-map supervision, false-negative-driven reweighting, and fold-wise physics-guided synthesis of positive CMBs and hard negatives. Experiments show improved lesion-level performance on VALDO Task 2 and strong recall and F1 on external AIBL SWI, supporting scalable CMB candidate extraction in large unlabeled cohorts.","CenSynCMB: Centre Maps and Physics-Guided Synthesis for Microbleed Detection  \narXiv :2607 .05325v 1 [ cs .CV] 6 Jul 2026  \nLucas He 1,2 , Hanyuan Zhang 1 , Krinos Li3 , Adama Fatima Saccoh4 Silvia Ingala5 , Rafael Rehwald6 , Marleen de Bruijne7 , Frederik Barkhof1,8,9  \nRhodri Davies2,10,†, Carole H. Sudre 1,2,11,†  \n1Hawkes Institute, University College London, UK 2Unit for Lifelong Health and Aging, University College London, UK 3Bioengineering Department and Imperial-X, Imperial College London, UK 4Institute of Cardiovascular Science, University College London, UK 5Department of Diagnostic Radiology, Copenhagen University Hospital, Denmark 6Neuroradiological Academic Unit, UCL Queen Square Institute of Neurology, London, UK 7Department of Radiology and Nuclear Medicine, Erasmus MC, The Netherlands 8Department of Radiology & Nuclear Medicine, Amsterdam UMC, Vrije Universiteit, The Netherlands 9Queen Square Institute of Neurology, University College London, UK 10Barts Heart Centre, St Bartholomew’s Hospital, London, UK 11Department of Biomedical Computing - School of Biomedical Engineering and Imaging Sciences, King’s College London, UK  \nAbstract—Cerebral microbleeds (CMBs) are MRI markers of small vessel disease and the microbleed component of amyloidrelated imaging abnormalities (ARIA-H), but their small size, sparsity, and similarity to vessels, calcification-like foci, and artefacts make automated detection difficult. We propose CenSynCMB, a centre-guided and mimic-aware framework combining a 3D Attention U-Net, auxiliary centre-map supervision, false-negative-driven reweighting, and fold-wise physics-guided synthesis of positive CMBs and labelled hard negatives. Synthetic data expose the detector to compact lesions and common mimics without validation or test leakage. On VALDO Task 2, CenSynCMB achieved the best local-comparison lesion-level F1 (74.3±8.8%, p = 0 .020); on external AIBL SWI, it achieved the highest local-comparison recall (88.5±6.9%, p = 0 .0058) and F1 (65.0±6.9%, p = 0 .0016). Together, these results support scalable CMB candidate extraction in large, unlabelled MRI cohorts, while highlighting cohort-specific calibration as the nextstep toward reliable burden estimation.  \nIndex Terms—cerebral microbleeds, MRI, multi-task learning, lesion detection, synthetic data  \nI. INTRODUCTION  \nA. Clinical Motivation  \nLarge studies assessing cerebral small vessel disease (CSVD) increasingly require more than visual summaries of lesion burden. CSVD is characterised on MRI by markers such as white matter hyperintensities (WMH), lacunes, enlarged perivascular spaces (EPVS), and cerebral microbleeds (CMBs) [2] . Among these markers, CMBs appear as small rounded hypointense foci on susceptibility-sensitive MRI such as T2∗ or susceptibility-weighted imaging (SWI), with diagnostic criteria commonly restricting them to lesions below 10 mm in diameter [3], [5] . They carry information about vascular brain injury, ageing, cognitive decline, and dementiarelated pathology [3] . Their anatomical distribution is clinically informative: lobar CMBs are more often associated with cerebral amyloid angiopathy, whereas deep CMBs are  \nCorresponding author: [Lucas.he.23@ucl.ac.uk](Lucas.he.23@ucl.ac.uk).  \nJoint senior authors.  \nmore often linked to hypertensive arteriopathy [3] . Accurate CMB quantification is therefore important not only for image interpretation and cohort studies, but also for monitoring amyloid-related imaging abnormalities with haemorrhage or haemosiderin deposition (ARIA-H) during anti-amyloid monoclonal antibody therapy [6] .  \nDespite this need, many CSVD studies still rely on visual assessment, including binary presence ratings, lesion counts, or structured scales such as the Microbleed Anatomical Rating Scale (MARS) and Brain Observer MicroBleed Scale (BOMBS) [4], [5] . These scales improve standardisation, but they still require trained readers and remain time-consuming for large datasets. The Vascular ","cbCaifw0NHzdgEVO","https://ap.wps.com/l/cbCaifw0NHzdgEVO","pdf",2842784,1,9,"English","en",105,"# Introduction\n## Clinical Motivation\n## Detection Challenges","[{\"question\":\"What makes cerebral microbleed (CMB) detection difficult in automated MRI analysis?\",\"answer\":\"CMBs are compact and spatially sparse, and their hypointense appearance can be confused with vessels in cross-section, calcification-like foci, susceptibility artefacts, and noise. Acquisition differences also change apparent lesion size, affecting transfer across sequences.\"},{\"question\":\"What core components does CenSynCMB use for improved detection?\",\"answer\":\"CenSynCMB combines a 3D Attention U-Net with auxiliary centre-map supervision and false-negative-driven reweighting. It also performs fold-wise physics-guided synthesis to generate positive CMBs and labelled hard negatives.\"},{\"question\":\"What performance improvements does CenSynCMB show on benchmark evaluations?\",\"answer\":\"On VALDO Task 2 it achieves the best local-comparison lesion-level F1 (74.3±8.8%, p=0.020). On external AIBL SWI it achieves the highest local-comparison recall (88.5±6.9%, p=0.0058) and F1 (65.0±6.9%, p=0.0016).\"}]",1784184100,23,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":27},"censyncmb-centre-maps-and-physics-guided-synthesis-for-microbleed-detection","",{"@graph":35,"@context":84},[36,53,67],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/censyncmb-centre-maps-and-physics-guided-synthesis-for-microbleed-detection/82935/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What makes cerebral microbleed (CMB) detection difficult in automated MRI analysis?","Question",{"text":74,"@type":75},"CMBs are compact and spatially sparse, and their hypointense appearance can be confused with vessels in cross-section, calcification-like foci, susceptibility artefacts, and noise. Acquisition differences also change apparent lesion size, affecting transfer across sequences.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"What core components does CenSynCMB use for improved detection?",{"text":79,"@type":75},"CenSynCMB combines a 3D Attention U-Net with auxiliary centre-map supervision and false-negative-driven reweighting. It also performs fold-wise physics-guided synthesis to generate positive CMBs and labelled hard negatives.",{"name":81,"@type":72,"acceptedAnswer":82},"What performance improvements does CenSynCMB show on benchmark evaluations?",{"text":83,"@type":75},"On VALDO Task 2 it achieves the best local-comparison lesion-level F1 (74.3±8.8%, p=0.020). On external AIBL SWI it achieves the highest local-comparison recall (88.5±6.9%, p=0.0058) and F1 (65.0±6.9%, p=0.0016).","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,119,122,126,129,133],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":21,"doc_module":4,"doc_module_name":45,"category_name":123,"show_sort_weight":124,"slug":125},"Religion & Spirituality",20,"religion-spirituality",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":127,"show_sort_weight":124,"slug":128},"World Cup","world-cup",{"id":130,"doc_module":4,"doc_module_name":45,"category_name":131,"show_sort_weight":130,"slug":132},10,"Lifestyle","lifestyle",{"id":134,"doc_module":4,"doc_module_name":45,"category_name":135,"show_sort_weight":105,"slug":136},19,"General","general"]