[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-327192-105":3,"doc-detail-327192-en":81,"detail-sidebar-cat-0-en-105":98},{"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","malariai-a-label-resilient-decoupled-framework-for-universal-cell-segmentation-and-explainable-stage-classification-in-dense-malaria-blood-smears","MalariAI - A Label-Resilient Decoupled Framework for Universal Cell Segmentation and Explainable Stage Classification in Dense Malaria Blood Smears","","Automated malaria diagnosis from blood smear microscopy is a critical challenge in global health AI, especially in resource-limited settings where expert microscopists are scarce. Existing deep learning systems fail reliably due to three compounding modes: missing annotations degrade recall, non-maximum suppression suppresses valid detections in dense overlaps, and whole-slide pipelines often lack per-cell spatial evidence for clinical audit. MalariAI introduces a unified two-stage decoupled framework addressing all three failure modes.",{"@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/healthcare/","Healthcare",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/malariai-a-label-resilient-decoupled-framework-for-universal-cell-segmentation-and-explainable-stage-classification-in-dense-malaria-blood-smears/327192/",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/malariai-a-label-resilient-decoupled-framework-for-universal-cell-segmentation-and-explainable-stage-classification-in-dense-malaria-blood-smears/327192.png","ImageObject",300,407,{"name":42,"@type":43},"Himbo","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-22","2026-09-21",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",7,{"@type":57,"mainEntity":58},"FAQPage",[59,65,69],{"name":60,"@type":61,"acceptedAnswer":62},"What are the three key failure modes MalariAI targets for clinical deployment?","Question",{"text":63,"@type":64},"It targets annotation incompleteness that lowers reliable recall, non-maximum suppression that suppresses valid detections in dense regions, and whole-slide pipelines that lack per-cell spatial evidence for clinical audit.","Answer",{"name":66,"@type":61,"acceptedAnswer":67},"How does MalariAI recover cells in dense, unannotated scenarios?",{"text":68,"@type":64},"Stage 1 uses an annotation-agnostic distance-transform guided watershed method to isolate every cell in a full blood smear image, recovering 75.95% of ground-truth cells by centroid localization without ground-truth input.",{"name":70,"@type":61,"acceptedAnswer":71},"How does MalariAI provide explainability at the instance level?",{"text":72,"@type":64},"Stage 2 generates Grad-CAM++ heatmaps per detected cell, offering instance-level spatial evidence so microscopists can verify predictions at the individual parasite level.","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},327192,1790181200,{"code":4,"msg":82,"data":83},"success",{"doc_id":79,"user_id":84,"nickname":42,"user_avatar":85,"doc_module":4,"category_id":55,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":86,"file_id":87,"file_url":88,"file_type":89,"file_size":90,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":91,"language":92,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":93,"faqs":94,"seo_title":95,"seo_description":12,"update_tm":96,"read_time":97},687197100911,"https://ap-avatar.wpscdn.com/avatar/a000239b6f1da00475?x-image-process=image/resize,m_fixed,w_180,h_180&k=1785132997149421697","arXiv :2607 .00385v1 [ ee ss .IV] 1 Jul 2026  \nHighlights  \nMalariAI: A Label-Resilient Decoupled Framework for Universal Cell Segmentation and Explainable Stage Classification in Dense Malaria Blood Smears  \nKaysarul Anas Apurba, Md Hasibul Hasan, Mohammed Ali, Tanzilur Rahman  \n• Annotation-agnostic watershed stage recovers 75 .95% of cells on BBBC041 without any ground-truth label.  \n• Decoupled design eliminates NMS, enabling reliable detection in dense, heavily-overlapping smear regions.  \n• EfficientNet-B0 with Focal Loss achieves 87 .5% and 75 .0% accuracy on rare schizont and gametocyte stages.  \n• Grad-CAM++ heatmaps are generated per detected cell, providing instance-level clinical explainability.  \n• First whole-slide malaria pipeline jointly addressing annotation resilience, density, and spatial XAI.  \nMalariAI: A Label-Resilient Decoupled Framework for Universal Cell Segmentation and Explainable Stage Classification in Dense Malaria Blood Smears  \nKaysarul Anas Apurbaa,∗, Md Hasibul Hasanb , Mohammed Alia , Tanzilur  \nRahmana  \na Department of Electrical and Computer Engineering, North South University, Dhaka  \n1229, Bangladesh  \nb Department of Computer Science and Engineering, International University of Business  \nAgriculture and Technology (IUBAT), Dhaka 1230, Bangladesh  \nAbstract  \nAutomated malaria diagnosis from blood smear microscopy is a critical challenge in global health AI; in resource-limited settings, the scarcity of expert microscopists remains the primary bottleneck to timely and accurate diagnosis. Three compounding failure modes prevent reliable clinical deployment of existing deep learning systems. First, end-to-end detectors treat unannotated cells as background during training, producing recall figures that are strongly influenced by annotation completeness rather than reflecting true cell recovery. Second, Non-Maximum Suppression tends to suppress valid detections in dense smear regions where infection counts matter most. Third, existing whole-slide detection pipelines lack per-cell spatial evidence for clinical audit, despite image-level explainability methods such as Grad-CAM having been applied to malaria image classification tasks. We present MalariAI, a  \n∗ Corresponding author.  \nEmail address: [kaysarulanas2@gmail.com](kaysarulanas2@gmail.com) (Kaysarul Anas Apurba) URL: [tanzilur.rahman@northsouth.edu](tanzilur.rahman@northsouth.edu) (Tanzilur Rahman)  \ntwo-stage decoupled framework that addresses all three failure modes in a unified pipeline. Stage 1 applies an annotation-agnostic distance-transform guided watershed algorithm to isolate every cell in a full 1600×1200 blood smear image, recovering 75.95% of ground-truth cells by centroid localisation across the 120-image NIH BBBC041 test set without any ground-truth input. Stage 2 fine-tunes EfficientNet-B0 with Focal Loss (γ = 2 .0 , per-class inversefrequency weights) on 64×64 crops, achieving 98.36% overall classification accuracy with 87.5% and 75.0% per-class accuracy on the rare schizont and gametocyte stages, compared to only 24.57% and 25.95% AP for a Faster R-CNN baseline on the same classes. Grad-CAM++ heatmaps generated per detected cell provide instance-level spatial evidence for clinical audit, enabling microscopists to verify model predictions at the individual parasite level without sacrificing classification performance.  \nKeywords: Malaria detection, Blood smear analysis, Instance segmentation, Watershed algorithm, EfficientNet, Focal Loss, Explainable AI,  \nGrad-CAM++, Medical image analysis, Decoupled framework  \n1. Introduction  \nMalaria remains one of the most consequential infectious diseases in contemporary global health. The World Health Organization reported 249 million cases worldwide in 2022, resulting in an estimated 608,000 deaths, of which 95% occurred in sub-Saharan Africa [1] . The burden falls disproportionatelyon children under five years of age and on populations with limited access to diagnostic in","cbCaitewPSv966oC","https://ap.wps.com/l/cbCaitewPSv966oC","pdf",15663999,75,"English","# Introduction\n## Clinical gold standard and bottleneck\n## Promise and limits of deep learning systems\n## Identified measurable failure modes","[{\"question\":\"What are the three key failure modes MalariAI targets for clinical deployment?\",\"answer\":\"It targets annotation incompleteness that lowers reliable recall, non-maximum suppression that suppresses valid detections in dense regions, and whole-slide pipelines that lack per-cell spatial evidence for clinical audit.\"},{\"question\":\"How does MalariAI recover cells in dense, unannotated scenarios?\",\"answer\":\"Stage 1 uses an annotation-agnostic distance-transform guided watershed method to isolate every cell in a full blood smear image, recovering 75.95% of ground-truth cells by centroid localization without ground-truth input.\"},{\"question\":\"How does MalariAI provide explainability at the instance level?\",\"answer\":\"Stage 2 generates Grad-CAM++ heatmaps per detected cell, offering instance-level spatial evidence so microscopists can verify predictions at the individual parasite level.\"}]","MalariAI - A Label-Resilient Decoupled Framework for Universal Cell Segmentation and Explainable Stage Classification in Dense Malaria Blood Smears | PDF",1789976646,189,{"code":4,"msg":82,"data":99},[100,104,108,112,117,122,125,130,135,138,142],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":101,"show_sort_weight":102,"slug":103},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":105,"show_sort_weight":106,"slug":107},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":109,"show_sort_weight":110,"slug":111},"Exam",70,"exam",{"id":113,"doc_module":4,"doc_module_name":25,"category_name":114,"show_sort_weight":115,"slug":116},5,"Comic",60,"comic",{"id":118,"doc_module":4,"doc_module_name":25,"category_name":119,"show_sort_weight":120,"slug":121},6,"Technology",50,"technology",{"id":55,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":123,"slug":124},40,"healthcare",{"id":126,"doc_module":4,"doc_module_name":25,"category_name":127,"show_sort_weight":128,"slug":129},8,"Research & Report",30,"research-report",{"id":131,"doc_module":4,"doc_module_name":25,"category_name":132,"show_sort_weight":133,"slug":134},9,"Religion & Spirituality",20,"religion-spirituality",{"id":133,"doc_module":4,"doc_module_name":25,"category_name":136,"show_sort_weight":133,"slug":137},"World Cup","world-cup",{"id":139,"doc_module":4,"doc_module_name":25,"category_name":140,"show_sort_weight":139,"slug":141},10,"Lifestyle","lifestyle",{"id":143,"doc_module":4,"doc_module_name":25,"category_name":144,"show_sort_weight":113,"slug":145},19,"General","general"]