[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-83937-en":3,"doc-seo-83937-105":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},83937,1099514068035,"Ezra","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",7,"Healthcare","Shape Over Intensity: Directional Topological Encoding for False Positive Reduction in Intracranial Aneurysm Detection","Automated intracranial aneurysm (IA) detection from CT angiography (CTA) is limited by high false-positive rates because conventional CNNs depend on local pixel intensities, confusing true saccular aneurysms with healthy vascular bifurcations. This issue is most severe for small, clinically critical lesions under 3 mm, where sensitivity often falls below 60%. A topology-aware plug-and-play framework evaluates Smooth Euler Characteristic Transform (SECT) versus persistence-based summaries on stratified RSNA 2025 subsets, yielding AUC 0.943, strong sub-3 mm performance, and scanner-agnostic robustness under leave-one-scanner-out validation.","arXiv :2607 .053 17v2 [ cs .CV] 7 Jul 2026  \nShape Over Intensity: Directional Topological Encoding for False Positive Reduction in Intracranial Aneurysm Detection  \nAkshay Gokhale [akshay.gokhale22@spit.ac.in](akshay.gokhale22@spit.ac.in)  \nComputer Science and Engineering Sardar Patel Institute of Technology Mumbai, India  \nMansi Dhamne [mansi.dhamne22@spit.ac.in](mansi.dhamne22@spit.ac.in)  \nComputer Engineering  \nSardar Patel Institute of Technology Mumbai, India  \nAbstract  \nAutomated detection of intracranial aneurysms (IAs) from CT angiography (CTA) is severely hindered by high false-positive rates. Traditional convolutional neural networks (CNNs) rely fundamentally on local pixel intensities, causing them to systematically confuse true saccular aneurysms with healthy vascular bifurcations. This geometric ambiguity is especially catastrophic for small, clinically critical lesions ( \u003C3 mm), where standard detection sensitivity often plummets below 60% .  \nTo resolve this, we propose a plug-and-play, topology-aware false-positive reduction framework. We evaluate the Smooth Euler Characteristic Transform (SECT)—a directional mathematical representation that encodes global 3D vascular geometry independently of intensity—against standard persistence-based summaries (Persistence Images and Landscapes) . The framework is rigorously stress-tested on a curated, heavily stratified subset of the multi-institutional RSNA 2025 dataset, explicitly designed to challenge models with anatomically plausible bifurcation mimics.  \nSECT demonstrates exceptional, classifier-agnostic discriminative power, achieving an AUC of 0.943, substantially outperforming direction-agnostic persistence methods (AUC ∼ 0.68) . Crucially, our topological filter exhibits a clinical performance inversion: it excelson the most challenging sub-3 mm cohort, maintaining a 0.943 AUC and 78.5% sensitivity even under a strict 95% specificity constraint. Furthermore, the representation proves highly resilient to hardware-specific artifacts, maintaining a 0.927 mean AUC under strict leave-one-scanner-out (LOGO) validation across four distinct manufacturers.  \nBy explicitly capturing asymmetric geometric invariants rather than intensity profiles, directional topological representations reliably resolve the primary structural confounderin IA detection. This establishes SECT as a highly robust, scanner-agnostic downstream filter ready for integration into hybrid deep-learning diagnostic pipelines.  \n1. Introduction  \nThe automated detection of Intracranial Aneurysms (IAs) remains a critical challenge in machine learning for healthcare. IAs are abnormal, localized dilations of cerebral arteries affecting approximately 3% to 7% of the adult population (Vlak et al. , 2011; Joo, 2025) . If left untreated, these weakened vessel walls can rupture, causing a fatal subarachnoid  \n© A. Gokhale & M. Dhamne.  \nShape Over Intensity  \nhemorrhage. While Digital Subtraction Angiography (DSA) remains the gold standard, Computed Tomography Angiography (CTA) is the primary non-invasive screening modality used in clinical practice (Hsu et al. , 2025) . However, reliably identifying IAs on CTAis notoriously difficult even for skilled radiologists. This difficulty is magnified for small lesions ( \u003C 3 mm in diameter), where sensitivity frequently drops to between 64% and 74%(Bizjak and ˇSpiclin, 2023) . Developing an automated diagnostic aid is paramount for early intervention, yet current deep learning systems generate an overwhelming number of false alarms, severely hindering their clinical viability.  \nAddressing this problem is challenging due to the inherent geometric limitations of standard convolutional neural networks (CNNs) . CTA natively suffers from lower spatial resolution and low vessel-to-background contrast, making small vascular structures visually indistinct because of surrounding tissue or noise. More fundamentally, modern CNNs rely heavily on local pixel intensities and text","cbCaiofhi9uJesS8","https://ap.wps.com/l/cbCaiofhi9uJesS8","pdf",5057815,1,36,"English","en",105,"# Abstract\n# Introduction","[{\"question\":\"为什么基于CTA的脑动脉瘤检测会产生较高的误报？\",\"answer\":\"传统CNN依赖局部像素强度与纹理统计，难以区分动脉瘤的囊状突起与健康血管分叉，因此会把相近的局部强度分布误判为动脉瘤，导致大量假阳性。\"},{\"question\":\"本文提出的核心方法是什么，用于降低假阳性？\",\"answer\":\"提出可即插即用的拓扑感知假阳性抑制框架，比较Smooth Euler Characteristic Transform（SECT）与基于持久性（Persistence）的表示（如Persistence Images和Landscapes）。SECT提供与强度无关的全局3D血管几何编码。\"},{\"question\":\"SECT在小于3 mm的病例上表现如何，且对不同扫描设备是否稳健？\",\"answer\":\"在严格95%特异度约束下，小于3 mm队列保持0.943 AUC，并实现78.5%的敏感度；同时在四家不同厂商的严格leave-one-scanner-out（LOGO）验证中，平均AUC达到0.927，显示出对硬件伪影的鲁棒性。\"}]",1784191546,91,{"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},"shape-over-intensity-directional-topological-encoding-for-false-positive-reduction-in-intracranial-aneurysm-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/healthcare/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/shape-over-intensity-directional-topological-encoding-for-false-positive-reduction-in-intracranial-aneurysm-detection/83937/",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},"为什么基于CTA的脑动脉瘤检测会产生较高的误报？","Question",{"text":74,"@type":75},"传统CNN依赖局部像素强度与纹理统计，难以区分动脉瘤的囊状突起与健康血管分叉，因此会把相近的局部强度分布误判为动脉瘤，导致大量假阳性。","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"本文提出的核心方法是什么，用于降低假阳性？",{"text":79,"@type":75},"提出可即插即用的拓扑感知假阳性抑制框架，比较Smooth Euler Characteristic Transform（SECT）与基于持久性（Persistence）的表示（如Persistence Images和Landscapes）。SECT提供与强度无关的全局3D血管几何编码。",{"name":81,"@type":72,"acceptedAnswer":82},"SECT在小于3 mm的病例上表现如何，且对不同扫描设备是否稳健？",{"text":83,"@type":75},"在严格95%特异度约束下，小于3 mm队列保持0.943 AUC，并实现78.5%的敏感度；同时在四家不同厂商的严格leave-one-scanner-out（LOGO）验证中，平均AUC达到0.927，显示出对硬件伪影的鲁棒性。","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},[92,96,100,104,109,114,117,122,127,130,134],{"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":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":115,"slug":116},40,"healthcare",{"id":118,"doc_module":4,"doc_module_name":45,"category_name":119,"show_sort_weight":120,"slug":121},8,"Research & Report",30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":105,"slug":137},19,"General","general"]