{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,24]],"date-time":"2026-10-24T23:41:10Z","timestamp":1792885270794,"version":"4.5.0"},"reference-count":28,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,11,25]],"date-time":"2025-11-25T00:00:00Z","timestamp":1764028800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Radiol."],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Aortic stenosis (AS) is diagnosed by echocardiography, the current gold standard, but examinations are often performed only after symptoms emerge, highlighting the need for earlier detection. Recently, artificial intelligence (AI)\u2013based screening using non-invasive and widely available modalities such as electrocardiography (ECG) and chest x-ray(CXR) has gained increasing attention for valvular heart disease. However, single-modality approaches have inherent limitations, and in clinical practice, multimodality assessment is common. In this study, we developed a multimodal AI model integrating ECG and CXR within a cooperative learning framework to evaluate its utility for earlier detection of AS.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>\n                      We retrospectively analyzed 23,886 patient records from 7,483 patients who underwent ECG, CXR, and echocardiography. A multimodal model was developed by combining a 1D ResNet50\u2013Transformer architecture for ECG data with an EfficientNet-based architecture for CXR. Cooperative learning was implemented using a loss function that allowed the ECG and CXR models to refine each other's predictions. We split the dataset into training, validation, and test sets, and performed 1,000 bootstrap iterations to assess model stability. AS was defined echocardiographically as peak velocity \u22652.5\u2005m\/s, mean pressure gradient \u226520\u2005mmHg, or aortic valve area \u22641.5\u2005cm\n                      <jats:sup>2<\/jats:sup>\n                      .\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Among 7,483 patients, 608 (8.1%) were diagnosed with AS. The multimodal model achieved a test AUROC of 0.812 (95% CI: 0.792\u20130.832), outperforming the ECG model (0.775, 95% CI: 0.753\u20130.796) and the CXR model (0.755, 95% CI: 0.732\u20130.777). Visualization techniques (Grad-CAM, Transformer attention) highlighted distinct yet complementary features in AS patients.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>The multimodal AI model via cooperative learning outperformed single-modality methods in AS detection and may aid earlier diagnosis and reduce clinical burden.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fradi.2025.1698680","type":"journal-article","created":{"date-parts":[[2025,11,25]],"date-time":"2025-11-25T06:25:16Z","timestamp":1764051916000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":6,"title":["Multimodal deep learning model for enhanced early detection of aortic stenosis integrating ECG and chest x-ray with cooperative learning"],"prefix":"10.3389","volume":"5","author":[{"given":"Shun","family":"Nagai","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Makoto","family":"Nishimori","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Masakazu","family":"Shinohara","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hidekazu","family":"Tanaka","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hiromasa","family":"Otake","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2025,11,25]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"e003287","DOI":"10.1161\/CIRCOUTCOMES.116.003287","article-title":"Quantifying the shift toward transcatheter aortic valve replacement in low-risk patients: a meta-analysis","volume":"10","author":"De Sciscio","year":"2017","journal-title":"Circ Cardiovasc Qual Outcomes"},{"key":"B2","doi-asserted-by":"publisher","first-page":"3515","DOI":"10.1093\/eurheartj\/ehw229","article-title":"Large-scale community echocardiographic screening reveals a major burden of undiagnosed valvular heart disease in older people: the oxvalve population cohort study","volume":"37","author":"d'Arcy","year":"2016","journal-title":"Eur Heart J"},{"key":"B3","doi-asserted-by":"publisher","first-page":"e72","DOI":"10.1161\/CIR.0000000000000923","article-title":"2020 ACC\/AHA guideline for the management of patients with valvular heart disease: a report of the American college of cardiology\/American heart association joint committee on clinical practice guidelines","volume":"143","author":"Otto","year":"2021","journal-title":"Circulation"},{"key":"B4","doi-asserted-by":"publisher","first-page":"524","DOI":"10.1016\/j.rec.2022.05.006","article-title":"2021 ESC\/EACTS guidelines for the management of valvular heart disease: developed by the task force for the management of valvular heart disease of the European society of cardiology (ESC) and the European association for cardio-thoracic surgery (EACTS)","volume":"75","author":"Vahanian","year":"2022","journal-title":"Rev Esp Cardiol (Engl Ed)"},{"key":"B5","doi-asserted-by":"publisher","first-page":"1289","DOI":"10.1002\/jum.16453","article-title":"Sound the alarm: the sonographer shortage is echoing across healthcare","volume":"43","author":"Won","year":"2024","journal-title":"J Ultrasound Med"},{"key":"B6","doi-asserted-by":"publisher","first-page":"e525","DOI":"10.1016\/S2589-7500(23)00107-3","article-title":"Artificial intelligence-based model to classify cardiac functions from chest radiographs: a multi-institutional, retrospective model development and validation study","volume":"5","author":"Ueda","year":"2023","journal-title":"Lancet Digit Health"},{"key":"B7","doi-asserted-by":"publisher","first-page":"613","DOI":"10.1016\/j.jacc.2022.05.029","article-title":"Deep learning electrocardiographic analysis for detection of left-sided valvular heart disease","volume":"80","author":"Elias","year":"2022","journal-title":"J Am Coll Cardiol"},{"key":"B8","doi-asserted-by":"publisher","first-page":"20","DOI":"10.1093\/ehjdh\/ztab102","article-title":"Artificial intelligence-based detection of aortic stenosis from chest radiographs","volume":"3","author":"Ueda","year":"2022","journal-title":"Eur Heart J Digit Health"},{"key":"B9","doi-asserted-by":"publisher","first-page":"765","DOI":"10.1161\/CIRCULATIONAHA.122.062646","article-title":"Detection of left ventricular systolic dysfunction from electrocardiographic images","volume":"148","author":"Sangha","year":"2023","journal-title":"Circulation"},{"key":"B10","doi-asserted-by":"publisher","first-page":"380","DOI":"10.1093\/ehjdh\/ztac033","article-title":"Multimodal deep learning enhances diagnostic precision in left ventricular hypertrophy","volume":"3","author":"Soto","year":"2022","journal-title":"Eur Heart J Digit Health"},{"key":"B11","doi-asserted-by":"publisher","first-page":"e2202113119","DOI":"10.1073\/pnas.2202113119","article-title":"Cooperative learning for multiview analysis","volume":"119","author":"Ding","year":"2022","journal-title":"Proc Natl Acad Sci USA"},{"key":"B12","doi-asserted-by":"publisher","first-page":"372","DOI":"10.1016\/j.echo.2017.02.009","article-title":"Recommendations on the echocardiographic assessment of aortic valve stenosis: a focused update from the European association of cardiovascular imaging and the American society of echocardiography","volume":"30","author":"Baumgartner","year":"2017","journal-title":"J Am Soc Echocardiogr"},{"key":"B13","doi-asserted-by":"publisher","first-page":"366","DOI":"10.1109\/TMI.2023.3305384","article-title":"Transformer-based spatio-temporal analysis for classification of aortic stenosis severity from echocardiography cine series","volume":"43","author":"Ahmadi","year":"2024","journal-title":"IEEE Trans Med Imaging"},{"key":"B14","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2016.90","article-title":"Deep residual learning for image recognition","author":"He","year":"2016"},{"key":"B15","doi-asserted-by":"publisher","first-page":"5998","DOI":"10.48550\/arXiv.1706.03762","article-title":"Attention is all you need","volume":"30","author":"Vaswani","year":"2017","journal-title":"Adv Neural Inf Process Syst"},{"key":"B16","article-title":"Adam: a method for stochastic optimization","author":"Kingma","year":"2014"},{"key":"B17","article-title":"Efficientnet: rethinking model scaling for convolutional neural networks","author":"Tan","year":""},{"key":"B18","doi-asserted-by":"publisher","first-page":"e607","DOI":"10.7717\/peerj-cs.607","article-title":"Chest x-ray pneumothorax segmentation using U-net with efficientnet and resnet architectures","volume":"7","author":"Abedalla","year":"2021","journal-title":"PeerJ Comput Sci"},{"key":"B19","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV.2017.74","article-title":"Grad-Cam: visual explanations from deep networks via gradient-based localization","author":"Selvaraju","year":"2017"},{"key":"B20","doi-asserted-by":"publisher","first-page":"2660","DOI":"10.1109\/TNNLS.2016.2599820","article-title":"Evaluating the visualization of what a deep neural network has learned","volume":"28","author":"Samek","year":"2017","journal-title":"IEEE Trans Neural Netw Learn Syst"},{"key":"B21","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV.2017.371","article-title":"Interpretable explanations of black boxes by meaningful perturbation","author":"Fong","year":"2017"},{"key":"B22","article-title":"Rise: randomized input sampling for explanation of black-box models","author":"Petsiuk","year":"2018"},{"key":"B23","doi-asserted-by":"publisher","first-page":"861","DOI":"10.1016\/j.patrec.2005.10.010","article-title":"An Introduction to ROC analysis","volume":"27","author":"Fawcett","year":"2006","journal-title":"Pattern Recognit Lett"},{"key":"B24","article-title":"Pytorch: an imperative style, high-performance deep learning library","volume":"32","author":"Paszke","year":"2019","journal-title":"Adv Neural Inf Process Syst"},{"key":"B25","doi-asserted-by":"publisher","first-page":"e033872","DOI":"10.1161\/JAHA.123.033872","article-title":"Moderate aortic valve stenosis is associated with increased mortality rate and lifetime loss: systematic review and meta-analysis of reconstructed time-to-event data of 409 680 patients","volume":"13","author":"Jacquemyn","year":"2024","journal-title":"J Am Heart Assoc"},{"key":"B26","doi-asserted-by":"publisher","first-page":"4592","DOI":"10.1093\/eurheartj\/ehad456","article-title":"Severe aortic stenosis detection by deep learning applied to echocardiography","volume":"44","author":"Holste","year":"2023","journal-title":"Eur Heart J"},{"key":"B27","doi-asserted-by":"publisher","first-page":"840262","DOI":"10.3389\/fcvm.2022.840262","article-title":"Use of multi-modal data and machine learning to improve cardiovascular disease care","volume":"9","author":"Amal","year":"2022","journal-title":"Front Cardiovasc Med"},{"key":"B28","doi-asserted-by":"publisher","first-page":"107058","DOI":"10.1016\/j.ultras.2023.107058","article-title":"Semi-supervised learning improves the performance of cardiac event detection in echocardiography","volume":"134","author":"Li","year":"2023","journal-title":"Ultrasonics"}],"container-title":["Frontiers in Radiology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fradi.2025.1698680\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,25]],"date-time":"2025-11-25T06:25:17Z","timestamp":1764051917000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fradi.2025.1698680\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11,25]]},"references-count":28,"alternative-id":["10.3389\/fradi.2025.1698680"],"URL":"https:\/\/doi.org\/10.3389\/fradi.2025.1698680","relation":{},"ISSN":["2673-8740"],"issn-type":[{"value":"2673-8740","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,11,25]]},"article-number":"1698680"}}