[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-83267-en":3,"doc-seo-83267-105":29,"detail-sidebar-cat-0-en-105":91},{"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":20,"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},83267,13056703019662,"Evangeline","https://ap-avatar.wpscdn.com/avatar/be000253a8e92610077?_k=1778726343310543188",8,"Research & Report","Cardiac MRI Through-Plane Super-Resolution Guided by Reference and Memory","Clinical cardiac MRI often uses high in-plane resolution but coarse through-plane resolution to reduce scan time under breath-hold and cardiac-motion constraints, which restricts reliable 3D analysis and diagnostic accuracy. STRMSR is a reference- and memory-guided through-plane super-resolution framework that reconstructs high-resolution 3D cardiac volumes by using subject-specific HR reference views and storing intermediate SR outputs in a memory bank. Coarse-to-fine contextual matching addresses spatial misalignment, while patch-wise dynamic feature aggregation produces content-adaptive fusion weights to suppress unreliable transfers and improve slice-to-slice consistency. Experiments on WHS demonstrate consistent gains at ×4 and ×8.","arXiv :2607 .0758 1v 1 [ cs .CV] 8 Jul 2026  \nCardiac MRI Through-Plane Super-Resolution Guided by Reference and Memory  \nShaoming Pan 1 , Chenchuhui Hu 1 , Leon Axel2 , and Meng Ye 1 ⋆  \n1 University of Texas At Arlington  \n2 New York University Grossman School of Medicine  \n[sxp0635@mavs.uta.edu](sxp0635@mavs.uta.edu)  \nAbstract. Clinical cardiac MRI is commonly acquired with high inplane resolution but coarse through-plane resolution to reduce scan time and accommodate breath-hold and cardiac-motion constraints, which limits 3D analysis and diagnostic accuracy. We propose STRMSR, a reference- and memory-guided through-plane super-resolution (SR) framework that reconstructs high-resolution (HR) cardiac volumes by leveraging HR reference views acquired from the same subject and intermediate SR results as the memory. Our method uses coarse-to-fine contextual matching to establish robust correspondence between lowresolution target and reference/memory images under spatial misalignment. A learnable patch-wise dynamic feature aggregation module predicts content-adaptive mixture weights for each local patch, effectively fusing dynamic information while suppressing unreliable feature transfers. The intermediate SR results stored in the memory bank ensure sliceto-slice consistency for the super-resolved 3D volume. Experiments on the WHS cardiac MRI dataset under two reference protocols, orthogonalplane views and long-axis chamber views, demonstrate consistent improvements over baselines at ×4 and ×8 upsampling factors.  \nKeywords: Super-resolution · Cardiac MRI · Matching · Memory  \n1 Introduction  \nCardiac MRI (CMR) relies on 2D breath-hold acquisitions with high in-plane resolution but coarse through-plane resolution, limiting downstream 3D analysis and diagnosis [21] . Standard CMR protocols routinely acquire a small set of long-axis views as complementary appearance cues for the short-axis stack or acquire orthogonal anisotropic 2D image stacks for 3D interpretation from different viewpoints. Recovering an isotropic 3D volume from such anisotropic stacks is commonly framed as slice-to-volume reconstruction (SVR) [12], which proceeds in two sequential steps: (i) slice alignment, which corrects rigid inter-breathhold misalignment between 2D image stacks via slice-to-slice or slice-to-volume registration [23, 1], and (ii) through-plane super-resolution (SR), which fuses the aligned anisotropic data into a coherent high-resolution (HR) 3D volume. The  \n⋆ Corresponding author  \n2 Shaoming Pan et al.  \nregistration step has been extensively studied, with mature solutions based on segmentation map intersection matching [20, 2, 24] . In this work, we focus on the second step and assume the input low-resolution (LR) stacks are already spatially aligned, aiming to learn an accurate through-plane SR model that recovers fine anatomical details using cross-view HR images as guidance.  \nEarly model-based methods formulate SR as an inverse problem regularized by handcrafted priors [19, 9, 18, 21], but rely on linear degradation assumptions and struggle to recover fine details at large upsampling factors. Deep learning has since become the dominant paradigm for MRI SR [11], with supervised methods learning LR-to-HR mapping via densely connected CNNs [4], multiscale convolutional networks [17, 6], and transformers [13] . However, most existing approaches operate on single-input or implicitly fuse multi-input data, limiting their ability to exploit complementary structural information from auxiliary views, and typically process slices independently, leading to inter-slice discontinuities in thereconstructed volume. Self-supervised methods such as SMORE [27] and SSGNN [22] avoid external training data but similarly lack mechanisms for crossview correspondence.  \nWe therefore propose a reference- and memory-based method, STRMSR, for the SR step of cardiac MRI SVR. Given a target LR volume as a video, we leverage auxiliary HR reference views from th","cbCaiqHZ2XHRUYVp","https://ap.wps.com/l/cbCaiqHZ2XHRUYVp","pdf",1026791,1,10,"English","en",105,"# Introduction\n## Slice alignment and through-plane SR in cardiac MRI\n## Related work: model-based, deep learning, and self-supervised SR\n# Method\n## Dual-Branch Transformer for feature extraction","[{\"question\":\"Why is through-plane resolution a limitation in clinical cardiac MRI?\",\"answer\":\"Clinical CMR typically provides high in-plane resolution but coarse through-plane resolution to meet scan-time, breath-hold, and cardiac-motion constraints, which limits accurate 3D analysis and diagnosis.\"},{\"question\":\"How does STRMSR use reference and memory to perform through-plane super-resolution?\",\"answer\":\"STRMSR reconstructs high-resolution 3D cardiac volumes using HR reference views from the same subject and stores intermediate SR outputs in a memory bank to propagate SR results along the third axis for better volumetric coherence.\"},{\"question\":\"What mechanisms enable robust detail transfer under geometric misalignment?\",\"answer\":\"STRMSR applies coarse-to-fine contextual matching to establish correspondence between low-resolution targets and reference/memory images despite spatial misalignment, and uses patch-wise dynamic feature aggregation to fuse dynamic information with content-adaptive mixture weights while suppressing unreliable transfers.\"}]",1784186404,25,{"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":86,"head_meta":88,"extra_data":90,"updated_unix":27},"cardiac-mri-through-plane-super-resolution-guided-by-reference-and-memory","",{"@graph":35,"@context":85},[36,53,68],{"@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/cardiac-mri-through-plane-super-resolution-guided-by-reference-and-memory/83267/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why is through-plane resolution a limitation in clinical cardiac MRI?","Question",{"text":75,"@type":76},"Clinical CMR typically provides high in-plane resolution but coarse through-plane resolution to meet scan-time, breath-hold, and cardiac-motion constraints, which limits accurate 3D analysis and diagnosis.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does STRMSR use reference and memory to perform through-plane super-resolution?",{"text":80,"@type":76},"STRMSR reconstructs high-resolution 3D cardiac volumes using HR reference views from the same subject and stores intermediate SR outputs in a memory bank to propagate SR results along the third axis for better volumetric coherence.",{"name":82,"@type":73,"acceptedAnswer":83},"What mechanisms enable robust detail transfer under geometric misalignment?",{"text":84,"@type":76},"STRMSR applies coarse-to-fine contextual matching to establish correspondence between low-resolution targets and reference/memory images despite spatial misalignment, and uses patch-wise dynamic feature aggregation to fuse dynamic information with content-adaptive mixture weights while suppressing unreliable transfers.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":21,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":21,"slug":133},"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]