[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-86433-en":3,"doc-seo-86433-105":30,"detail-sidebar-cat-0-en-105":92},{"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":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},86433,137441390410,"Hazel","https://ap-avatar.wpscdn.com/avatar/2000252f4ab5702993?_k=1776741390130283984",8,"Research & Report","ECG-LDC A Hardware Efficient Low Dimensional Computing Framework for ECG Arrhythmia Classification","Continuous cardiac monitoring on wearable devices requires ECG arrhythmia classifiers that are accurate, energy-efficient, and feasible on resource-constrained hardware. While deep neural models achieve strong ECG performance, their large parameter counts and heavy multiply-accumulate workloads hinder deployment on low-cost edge platforms. ECG-LDC proposes a hardware-software co-design adapting Low-Dimensional Computing with dual encoders and dedicated value/feature codebooks for morphology and RR-interval dynamics. Built with binary representations on Pynq-Z2, it reaches 97.18% accuracy with 3.86 kB memory, offering major memory savings over TinyML at modest accuracy loss.","ECG-LDC: A Hardware-Efficient Low-Dimensional Computing Framework for ECG Arrhythmia Classification  \nAnh Tran, Khanh Tran, Cuong Do  \narXiv :2607 .09680v1 [ ee ss . SP] 13 Jun 2026  \nAbstract—Continuous cardiac monitoring in wearable devices demands classifiers that are simultaneously accurate, energyefficient, and deployable on resource-constrained hardware. While deep neural network approaches have demonstrated high classification accuracy for electrocardiogram (ECG) arrhythmia detection, their substantial parameter counts and reliance on multiply-accumulate-intensive operations make them impractical for low-cost edge platforms. In this work, we propose ECG-LDC, a hardware-software co-design framework that adapts LowDimensional Computing (LDC) for real-time ECG arrhythmia classification. ECG-LDC employs a dual-encoder architecture with dedicated value and feature codebooks to independently encode morphological waveform features and RR-interval temporal features, enabling effective capture of both intra-beat and inter-beat cardiac dynamics. The framework encompasses data preprocessing, model training, and a hardware accelerator architecture prototyped on the Pynq-Z2 platform. Implemented using binary representations and XOR/XNOR-based operations, ECG-LDC achieves 97. 18% accuracy with a memory footprint of only 3.86 kB. ECG-LDC sacrifices approximately 1.8% accuracy versus SOTA TinyML classifiers but achieves 11–570 × reduction in memory usage; among FPGA-based five-class arrhythmia classifiers, it delivers the highest accuracy with up to 2.4 × fewer LUTs and zero DSP block utilization, affirming its suitability for real-time arrhythmia detection on resource-constrained wearable platforms.  \nIndex Terms—TinyML, edge computing, FPGA, hardware acceleration, arrhythmia classification, wearable devices.  \nI. INTRODUCTION  \nCARDIOVASCULAR diseases (CVDs) are a group of  \ndisorders affecting the heart and blood vessels, encompassing conditions such as coronary artery disease, heart failure, and stroke. CVDs remain the leading cause of mortality worldwide, accounting for an estimated 19.2 million deaths annually [1] . A critical component of CVD treatment and clinical management is the electrocardiogram (ECG), a noninvasive tool that records the heart’s electrical activity using electrodes placed on the body surface. The ECG captures subtle variations in cardiac electrical signals that reflect different  \nManuscript created June, 2026 . This work was supported by the VinUni– Illinois Smart Health Center, VinUniversity, Hanoi 100000, Vietnam,, under Project No. 10000167, titled “Point of Care and Telehealth Diagnostics for Data-Driven Smart Health Systems.” (Corresponding authors: Anh Tran and Cuong Do.)  \nAnh Tran is with the School of Engineering and Applied Science, University of Pennsylvania, Philadelphia, PA 19104, USA, and also with the VinUni– Illinois Smart Health Center, VinUniversity, Hanoi 100000, Vietnam (email: [anh.thh@vinuni.edu.vn](anh.thh@vinuni.edu.vn)) .  \nKhanh Tran is with the College of Engineering and Computer Science, VinUniversity, Hanoi 100000, Vietnam (e-mail: [khanh.tg@vinuni.edu.vn](khanh.tg@vinuni.edu.vn)).  \nCuong Do is with the College of Engineering and Computer Science, VinUniversity, Hanoi 100000, Vietnam, and also with the VinUni– Illinois Smart Health Center, VinUniversity, Hanoi 100000, Vietnam (e-mail: [cuong.dd@vinuni.edu.vn](cuong.dd@vinuni.edu.vn)).  \nheart conditions. It provides rich, insightful information for accurate detection of arrhythmias, i.e., abnormal heartbeats, facilitating early diagnosis of potential cardiac diseases.  \nIn recent years, deep neural networks have emerged as effective solutions for ECG signal analysis. Various approaches leveraging residual connections, convolutional architectures, and attention mechanisms have been proposed, with works such as Allam et al. [2] and Islam et al. [3] demonstrating classification accuracies exceeding 99% on the MIT-BIH Arrhythmia D","cbCaimXGeOxQjYWh","https://ap.wps.com/l/cbCaimXGeOxQjYWh","pdf",463034,5,1,10,"English","en",105,"# Introduction\n## Cardiovascular disease and the role of ECG\n## Limits of deep neural networks for edge deployment\n## Hardware accelerators and FPGA trade-offs\n## Low-Dimensional Computing (LDC) overview\n# Motivation","[{\"question\":\"Why are conventional deep neural networks difficult to deploy on low-cost wearable ECG platforms?\",\"answer\":\"They typically have large parameter counts and require many multiply-accumulate operations and substantial memory, which exceed the capacity of resource-constrained edge devices.\"},{\"question\":\"What is the core idea of ECG-LDC for ECG arrhythmia classification?\",\"answer\":\"ECG-LDC adapts Low-Dimensional Computing using a dual-encoder architecture with separate value and feature codebooks to encode waveform morphology features and RR-interval temporal features.\"},{\"question\":\"What hardware and implementation approach does ECG-LDC use, and what results does it achieve?\",\"answer\":\"The framework is prototyped on the Pynq-Z2 platform using binary representations and XOR/XNOR-based operations, achieving 97.18% accuracy with a 3.86 kB memory footprint.\"}]",1784211717,25,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":28},"ecg-ldc-a-hardware-efficient-low-dimensional-computing-framework-for-ecg-arrhythmia-classification","",{"@graph":36,"@context":86},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/ecg-ldc-a-hardware-efficient-low-dimensional-computing-framework-for-ecg-arrhythmia-classification/86433/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":24,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-07-27","2026-07-16",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"Why are conventional deep neural networks difficult to deploy on low-cost wearable ECG platforms?","Question",{"text":76,"@type":77},"They typically have large parameter counts and require many multiply-accumulate operations and substantial memory, which exceed the capacity of resource-constrained edge devices.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"What is the core idea of ECG-LDC for ECG arrhythmia classification?",{"text":81,"@type":77},"ECG-LDC adapts Low-Dimensional Computing using a dual-encoder architecture with separate value and feature codebooks to encode waveform morphology features and RR-interval temporal features.",{"name":83,"@type":74,"acceptedAnswer":84},"What hardware and implementation approach does ECG-LDC use, and what results does it achieve?",{"text":85,"@type":77},"The framework is prototyped on the Pynq-Z2 platform using binary representations and XOR/XNOR-based operations, achieving 97.18% accuracy with a 3.86 kB memory footprint.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":93},[94,98,102,106,110,115,120,123,128,131,134],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":20,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":22,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":22,"slug":133},"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":20,"slug":137},19,"General","general"]