[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-345276-105":3,"detail-sidebar-cat-0-en-105":81,"doc-detail-345276-en":130},{"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","ensemble-transformer-based-neural-networks-detect-heart-murmur-in-phonocardiogram-recordings","Ensemble Transformer-Based Neural Networks Detect Heart Murmur In Phonocardiogram Recordings","","Cardiac auscultation via phonocardiogram (PCG) remains widely used for assessing heart mechanical function in congenital heart disease, but clinical interpretation is limited by the need for expert knowledge. This study proposes ensemble transformer-based neural networks to detect heart murmurs from PCG recordings and predict clinical outcomes for the George B. Moody PhysioNet 2022 Challenge. Raw PCG signals are converted into wavelet power features for deep learning. The approach achieves 0.855 training and 0.761 hidden-validation accuracy for murmur detection, with favorable outcome costs and strong competition ranks.",{"@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/research-report/","Research & Report",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/ensemble-transformer-based-neural-networks-detect-heart-murmur-in-phonocardiogram-recordings/345276/",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/ensemble-transformer-based-neural-networks-detect-heart-murmur-in-phonocardiogram-recordings/345276.png","ImageObject",300,407,{"name":42,"@type":43},"Aria","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-29","2026-09-22",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",6,{"@type":57,"mainEntity":58},"FAQPage",[59,65,69],{"name":60,"@type":61,"acceptedAnswer":62},"What problem does the study address in PCG-based congenital heart disease assessment?","Question",{"text":63,"@type":64},"The study targets limited diagnostic sensitivity caused by the need for experienced practitioners to interpret phonocardiogram signals.","Answer",{"name":66,"@type":61,"acceptedAnswer":67},"How are the PCG recordings transformed before entering the deep learning models?",{"text":68,"@type":64},"The method selects the first 40 seconds, applies z-score normalization, then performs wavelet transform decomposition (Symlets 8) and computes power and derived features to form a multi-feature representation.",{"name":70,"@type":61,"acceptedAnswer":71},"How does the proposed model evaluate performance for murmur detection and clinical outcomes?",{"text":72,"@type":64},"It reports maximum murmur detection accuracy of 0.855 on training and 0.761 on hidden validation, and provides overall clinical outcome cost values for both datasets, along with ranking positions in the challenge.","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},345276,1791020440,{"code":4,"msg":82,"data":83},"success",[84,88,92,96,101,105,110,114,119,122,126],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":85,"show_sort_weight":86,"slug":87},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":89,"show_sort_weight":90,"slug":91},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":93,"show_sort_weight":94,"slug":95},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":25,"category_name":98,"show_sort_weight":99,"slug":100},5,"Comic",60,"comic",{"id":55,"doc_module":4,"doc_module_name":25,"category_name":102,"show_sort_weight":103,"slug":104},"Technology",50,"technology",{"id":106,"doc_module":4,"doc_module_name":25,"category_name":107,"show_sort_weight":108,"slug":109},7,"Healthcare",40,"healthcare",{"id":111,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":112,"slug":113},8,30,"research-report",{"id":115,"doc_module":4,"doc_module_name":25,"category_name":116,"show_sort_weight":117,"slug":118},9,"Religion & Spirituality",20,"religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":25,"category_name":120,"show_sort_weight":117,"slug":121},"World Cup","world-cup",{"id":123,"doc_module":4,"doc_module_name":25,"category_name":124,"show_sort_weight":123,"slug":125},10,"Lifestyle","lifestyle",{"id":127,"doc_module":4,"doc_module_name":25,"category_name":128,"show_sort_weight":97,"slug":129},19,"General","general",{"code":4,"msg":82,"data":131},{"doc_id":79,"user_id":132,"nickname":42,"user_avatar":133,"doc_module":4,"category_id":111,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":33,"language":139,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":140,"faqs":141,"seo_title":142,"seo_description":12,"update_tm":143,"read_time":123,"preview_page_count":33,"preview_urls":144},2336464648322,"https://ap-avatar.wpscdn.com/avatar/2200025388227c56fec?_k=1778556882303663488","Ensemble Transformer-Based Neural Networks Detect Heart Murmur In  \nPhonocardiogram Recordings  \nMohanad Alkhodari 1 , Syafiq Kamarul Azman2 , Leontios J. Hadjileontiadis 1,3 , Ahsan H. Khandoker 1  \n1Healthcare Engineering Innovation Center (HEIC), Department of Biomedical Engineering, Khalifa University, Abu Dhabi, United Arab Emirates  \n2AIQ, ADNOC H.Q., Abu Dhabi, United Arab Emirates  \n3Department of Electrical and Computer Engineering, Aristotle University of Thessaloniki, Greece  \nAbstract  \nCardiac auscultation through phonocardiogram (PCG) is still the most commonly used approach for evaluating the mechanical functionality of the heart when diagnosing congenital heart disease. Despite of its time- and cost-effectiveness, it is still limited due to the extensive need for clinical expertise for interpretation. In this study, we propose the use of ensemble transformer-based neural networks to aid in the detection of heart murmur in PCG recordings and for the prediction of clinical outcomes of patients as part of George B. Moody PhysioNet 2022 Challenge. Our team, Care4MyHeart, developed an approach that transforms the raw PCG recordings into wavelet power features signals for the use within the proposed deep learning models. We have achieved a maximum accuracy of 0.855 and 0. 761 for murmur detection in the training and hidden validation datasets, respectively. In addition, we had an overall clinical outcome cost of 9980 and 11490 for both datasets, respectively. Our team was ranked 12th/305 for murmur detection and 173rd/305 for clinical outcome predictions. We had the lowest clinical outcome cost on the validation set (9737, 41st/305) with a murmur detection score of 0.730 when reducing the number of features used to train the models.  \n1. Introduction  \nCongenital heart disease is a birth defect in the structure of the heart that affects nearly 1 .2% of newborns worldwide [1] . Early assessment through non-invasive cardiac auscultation devices, i.e., phonocardiogram (PCG), can unveil fundamental information on the mechanical malfunctioning caused by this disease [2, 3] . However, although PCG is a time-and cost-effective tool, its technology still lacks proper interpretation owing to the need of experienced practitioners and thus, it has a limited diagnostic sensitivity [4] .  \nTo address this concern, we propose a simple, yet effective, deep learning approach based on attention transformers to detect cardiac malfunction in PCG recordings, which are usually represented as heart murmurs or abnormal sound patterns. We used patient data from The George B. Moody PhysioNet 2022 Challenge [5, 6] with an objective to provide a prediction to whether a patient has absent, present, or unknown murmurs plus identifying the clinical outcomes as either abnormal or normal.  \n2. Methods  \n2.1. Data preparation  \nThe challenge data included 1568 patients from the pediatric population, out of which 942 were released as training while the remaining were hidden for validation/testing. Each patient had one or more PCG recordings from four auscultation locations, i.e., aortic valve (AV), mitral valve (MV), pulmonary valve (PV), tricuspid valve (TV), and other (Phc) . All recordings were taken sequentially and not simultaneously.  \nWe started our approach by arranging all patient data to include four channels arranged as AV, MV, PV, and TV. Ifa patient had less than four channels, the previous channel was duplicated once or multiple times as needed. Then, we selected the first 40 seconds from each recording. Many signals were recorded for shorter duration, therefore, we padded them by duplicating the same signal to reach up to 40 seconds. At the end, we apply z-score normalization for the four-channel recording.  \n2.2. Transformation to wavelet features  \nInitially, we transformed each PCG signal into wavelet transform-based power features (Fig. 1) . For each signal, we used a 32-sample wide sliding window with a shifting interval of 32 sam","cbCaikHQQytRgxDh","https://ap.wps.com/l/cbCaikHQQytRgxDh","pdf",634823,"English","# Abstract\n# Introduction\n# Methods\n## Data preparation\n## Transformation to wavelet features\n## Transformer-based neural network","[{\"question\":\"What problem does the study address in PCG-based congenital heart disease assessment?\",\"answer\":\"The study targets limited diagnostic sensitivity caused by the need for experienced practitioners to interpret phonocardiogram signals.\"},{\"question\":\"How are the PCG recordings transformed before entering the deep learning models?\",\"answer\":\"The method selects the first 40 seconds, applies z-score normalization, then performs wavelet transform decomposition (Symlets 8) and computes power and derived features to form a multi-feature representation.\"},{\"question\":\"How does the proposed model evaluate performance for murmur detection and clinical outcomes?\",\"answer\":\"It reports maximum murmur detection accuracy of 0.855 on training and 0.761 on hidden validation, and provides overall clinical outcome cost values for both datasets, along with ranking positions in the challenge.\"}]","Ensemble Transformer-Based Neural Networks Detect Heart Murmur In Phonocardiogram Recordings | PDF",1790057095,[145,149,153,157],{"page_no":22,"url":146,"width":147,"height":148},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/70bcc986b632006f464069c05417d575.webp",983,1272,{"page_no":26,"url":150,"width":151,"height":152},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/27c617cc56921d752d8b334d7a44fec0.webp",982,1270,{"page_no":30,"url":154,"width":155,"height":156},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/a7b977e3ae3b5b5a7b9048ba66cc5102.webp",981,1269,{"page_no":33,"url":158,"width":159,"height":160},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/4ebbb9111298b8a47e79562070af35a6.webp",980,1268]