{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,14]],"date-time":"2026-11-14T13:46:59Z","timestamp":1794664019782,"version":"4.9.1"},"reference-count":32,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T00:00:00Z","timestamp":1780272000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T00:00:00Z","timestamp":1780272000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["npj Digit. Med."],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Artificial intelligence (AI) has shown strong performance in cardiology, but most approaches rely on sensing modalities whose physical limitations constrain available information. Magnetocardiography (MCG) records the cardiac magnetic field with less tissue distortion than surface potentials and may preserve higher-dimensional spatiotemporal electrophysiological structure. Here, we investigated whether combining MCG with self-supervised learning enables physiologically meaningful cardiac representations. We developed MCG2Vec, a contrastive encoder trained directly on raw 64-channel MCG recordings. Using recordings from 1732 consecutive patients, learned embeddings were evaluated with task-specific probes for multivessel coronary artery disease, reduced left ventricular ejection fraction, and paroxysmal atrial fibrillation risk from sinus-rhythm recordings. The representations enabled discrimination of multivessel coronary artery disease (area under the receiver operating characteristic curve (AUC) 0.89), reduced left ventricular ejection fraction (AUC 0.81), and atrial fibrillation risk (AUC 0.77). Attribution analyses revealed probe-specific temporal and spatial patterns corresponding to ventricular depolarization, repolarization, atrial activation dynamics, and coronary territories, supporting physiological interpretability. These findings suggest that higher-fidelity sensing combined with self-supervised representation learning can yield structured and explainable embeddings from non-invasive cardiac magnetic field recordings. More broadly, the study highlights measurement physics as an important determinant of what medical AI systems can learn.<\/jats:p>","DOI":"10.1038\/s41746-026-02819-8","type":"journal-article","created":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T04:03:47Z","timestamp":1780286627000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":3,"title":["Self-supervised representation learning reveals explainable physiological structure in high-dimensional magnetocardiography"],"prefix":"10.1038","volume":"9","author":[{"given":"Dominik D.","family":"Kranz","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Oru\u00e7","family":"Kahriman","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dominic","family":"Dischl","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sascha","family":"Treskatsch","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Andr\u00e9","family":"Sander","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Johannes","family":"Brachmann","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jai-Wun","family":"Park","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Niels","family":"Wessel","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,6,1]]},"reference":[{"key":"2819_CR1","doi-asserted-by":"publisher","unstructured":"Stark, B. A. et al. Global, Regional, and National Burden of Cardiovascular Diseases and Risk Factors in 204 Countries and Territories, 1990\u20132023. JACC S0735109725074285 (2025) https:\/\/doi.org\/10.1016\/j.jacc.2025.08.015.","DOI":"10.1016\/j.jacc.2025.08.015"},{"key":"2819_CR2","doi-asserted-by":"publisher","first-page":"e89","DOI":"10.1161\/CIR.0000000000001258","volume":"150","author":"DS Kazi","year":"2024","unstructured":"Kazi, D. S. et al. Forecasting the economic burden of cardiovascular disease and stroke in the United States through 2050: a presidential advisory from the American Heart Association. Circulation 150, e89\u2013e101 (2024).","journal-title":"Circulation"},{"key":"2819_CR3","doi-asserted-by":"publisher","first-page":"2467","DOI":"10.1016\/j.jacc.2016.08.069","volume":"68","author":"A Arbab-Zadeh","year":"2016","unstructured":"Arbab-Zadeh, A. & Fuster, V. The risk Continuum of Atherosclerosis and its implications for defining CHD by coronary angiography. J. Am. Coll. Cardiol. 68, 2467\u20132478 (2016).","journal-title":"J. Am. Coll. Cardiol."},{"key":"2819_CR4","doi-asserted-by":"publisher","first-page":"407","DOI":"10.1093\/eurheartj\/ehz425","volume":"41","author":"J Knuuti","year":"2020","unstructured":"Knuuti, J. et al. 2019 ESC Guidelines for the diagnosis and management of chronic coronary syndromes. Eur. Heart J. 41, 407\u2013477 (2020).","journal-title":"Eur. Heart J."},{"key":"2819_CR5","doi-asserted-by":"publisher","first-page":"1306","DOI":"10.1161\/CIRCULATIONAHA.106.180200","volume":"115","author":"P Kligfield","year":"2007","unstructured":"Kligfield, P. et al. Recommendations for the Standardization and Interpretation of the Electrocardiogram: Part I: The Electrocardiogram and Its Technology: A Scientific Statement From the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society Endorsed by the International Society for Computerized Electrocardiology. Circulation 115, 1306\u20131324 (2007).","journal-title":"Circulation"},{"key":"2819_CR6","doi-asserted-by":"publisher","first-page":"1447","DOI":"10.1093\/europace\/euaa165","volume":"22","author":"H Pereira","year":"2020","unstructured":"Pereira, H., Niederer, S. & Rinaldi, C. A. Electrocardiographic imaging for cardiac arrhythmias and resynchronization therapy. Europace 22, 1447\u20131462 (2020).","journal-title":"Europace"},{"key":"2819_CR7","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1007\/s12350-011-9490-9","volume":"19","author":"TM Bateman","year":"2012","unstructured":"Bateman, T. M. Advantages and disadvantages of PET and SPECT in a busy clinical practice. J. Nucl. Cardiol. 19, 3\u201311 (2012).","journal-title":"J. Nucl. Cardiol."},{"key":"2819_CR8","doi-asserted-by":"publisher","DOI":"10.1186\/s12968-020-00607-1","volume":"22","author":"CM Kramer","year":"2020","unstructured":"Kramer, C. M. et al. Standardized cardiovascular magnetic resonance imaging (CMR) protocols: 2020 update. J. Cardiovasc. Magn. Reson. 22, 17 (2020).","journal-title":"J. Cardiovasc. Magn. Reson."},{"key":"2819_CR9","doi-asserted-by":"publisher","first-page":"413","DOI":"10.1103\/RevModPhys.65.413","volume":"65","author":"M H\u00e4m\u00e4l\u00e4inen","year":"1993","unstructured":"H\u00e4m\u00e4l\u00e4inen, M., Hari, R., Ilmoniemi, R. J., Knuutila, J. & Lounasmaa, O. V. Magnetoencephalography\u2014theory, instrumentation, and applications to noninvasive studies of the working human brain. Rev. Mod. Phys. 65, 413\u2013497 (1993).","journal-title":"Rev. Mod. Phys."},{"key":"2819_CR10","doi-asserted-by":"publisher","first-page":"291","DOI":"10.1586\/14737159.5.3.291","volume":"5","author":"R Fenici","year":"2005","unstructured":"Fenici, R., Brisinda, D. & Meloni, A. M. Clinical application of magnetocardiography. Expert Rev. Mol. Diagn. 5, 291\u2013313 (2005).","journal-title":"Expert Rev. Mol. Diagn."},{"key":"2819_CR11","doi-asserted-by":"publisher","first-page":"474","DOI":"10.1016\/S0013-4694(97)00043-6","volume":"103","author":"YC Okada","year":"1997","unstructured":"Okada, Y. C., Wu, J. & Kyuhou, S. Genesis of MEG signals in a mammalian CNS structure. Electroencephalogr. Clin. Neurophysiol. 103, 474\u2013485 (1997).","journal-title":"Electroencephalogr. Clin. Neurophysiol."},{"key":"2819_CR12","unstructured":"Baevski, A., Zhou, H., Mohamed, A. & Auli, M. wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations. In Advances in Neural Information Processing Systems 33 (2020)"},{"key":"2819_CR13","unstructured":"Chen, T., Kornblith, S., Swersky, K., Norouzi, M. & Hinton, G. Big Self-Supervised Models are Strong Semi-Supervised Learners. In Advances in Neural Information Processing Systems (NeurIPS, 2020)."},{"key":"2819_CR14","unstructured":"Xiao, Q. et al. BrainOmni: A Brain Foundation Model for Unified EEG and MEG Signals. In Advances in Neural Information Processing Systems (NeurIPS, 2025)."},{"key":"2819_CR15","doi-asserted-by":"publisher","first-page":"108420","DOI":"10.1016\/j.bspc.2025.108420","volume":"112","author":"K Ma","year":"2026","unstructured":"Ma, K., Zhang, T., Zhang, H. & Huang, W. Self-supervised contrastive learning achieves 12-lead ECG classification. Biomed. Signal Process. Control 112, 108420 (2026).","journal-title":"Biomed. Signal Process. Control"},{"key":"2819_CR16","doi-asserted-by":"crossref","unstructured":"Yang, C., Westover, M. & Sun, J. BIOT: Biosignal Transformer for Cross-data Learning in the Wild. In Advances in Neural Information Processing Systems (eds Oh, A. et al.) vol.36 78240\u201378260 (Curran Associates, Inc., 2023).","DOI":"10.52202\/075280-3420"},{"key":"2819_CR17","doi-asserted-by":"publisher","unstructured":"Selvaraju, R. R. et al. Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization. In 2017 IEEE International Conference on Computer Vision (ICCV) 618\u2013626 (IEEE, Venice, 2017). https:\/\/doi.org\/10.1109\/ICCV.2017.74.","DOI":"10.1109\/ICCV.2017.74"},{"key":"2819_CR18","doi-asserted-by":"publisher","first-page":"65","DOI":"10.1038\/s41591-018-0268-3","volume":"25","author":"AY Hannun","year":"2019","unstructured":"Hannun, A. Y. et al. Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network. Nat. Med. 25, 65\u201369 (2019).","journal-title":"Nat. Med."},{"key":"2819_CR19","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-020-15432-4","volume":"11","author":"AH Ribeiro","year":"2020","unstructured":"Ribeiro, A. H. et al. Automatic diagnosis of the 12-lead ECG using a deep neural network. Nat. Commun. 11, 1760 (2020).","journal-title":"Nat. Commun."},{"key":"2819_CR20","doi-asserted-by":"publisher","first-page":"221","DOI":"10.1038\/s41586-025-09227-0","volume":"644","author":"TJ Poterucha","year":"2025","unstructured":"Poterucha, T. J. et al. Detecting structural heart disease from electrocardiograms using AI. Nature 644, 221\u2013230 (2025).","journal-title":"Nature"},{"key":"2819_CR21","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-025-14054-4","volume":"15","author":"D Dischl","year":"2025","unstructured":"Dischl, D. et al. Magnetoionography enhances diagnostic accuracy of magnetocardiography in coronary artery disease. Sci. Rep. 15, 31289 (2025).","journal-title":"Sci. Rep."},{"key":"2819_CR22","doi-asserted-by":"publisher","first-page":"1444963","DOI":"10.3389\/fcvm.2024.1444963","volume":"11","author":"J-W Park","year":"2024","unstructured":"Park, J.-W. et al. Editorial: Current proceedings in magnetocardiology-past, present, future. Front. Cardiovasc. Med. 11, 1444963 (2024).","journal-title":"Front. Cardiovasc. Med."},{"key":"2819_CR23","doi-asserted-by":"publisher","DOI":"10.1038\/s41698-024-00553-6","volume":"8","author":"F Markowetz","year":"2024","unstructured":"Markowetz, F. All models are wrong and yours are useless: making clinical prediction models impactful for patients. npj Precis. Oncol. 8, 54 (2024).","journal-title":"npj Precis. Oncol."},{"key":"2819_CR24","doi-asserted-by":"publisher","DOI":"10.1038\/s41746-024-01118-4","volume":"7","author":"C Ong Ly","year":"2024","unstructured":"Ong Ly, C. et al. Shortcut learning in medical AI hinders generalization: method for estimating AI model generalization without external data. npj Digit. Med. 7, 124 (2024).","journal-title":"npj Digit. Med."},{"key":"2819_CR25","doi-asserted-by":"publisher","first-page":"842","DOI":"10.1016\/j.jacr.2023.06.025","volume":"20","author":"I Banerjee","year":"2023","unstructured":"Banerjee, I. et al. Shortcuts\u2019 causing bias in radiology artificial intelligence: causes, evaluation, and mitigation. J. Am. Coll. Radio. 20, 842\u2013851 (2023).","journal-title":"J. Am. Coll. Radio."},{"key":"2819_CR26","doi-asserted-by":"publisher","first-page":"719","DOI":"10.1038\/s41551-023-01056-8","volume":"7","author":"RJ Chen","year":"2023","unstructured":"Chen, R. J. et al. Algorithmic fairness in artificial intelligence for medicine and healthcare. Nat. Biomed. Eng. 7, 719\u2013742 (2023).","journal-title":"Nat. Biomed. Eng."},{"key":"2819_CR27","doi-asserted-by":"publisher","first-page":"e1002683","DOI":"10.1371\/journal.pmed.1002683","volume":"15","author":"JR Zech","year":"2018","unstructured":"Zech, J. R. et al. Variable generalization performance of a deep learning model to detect pneumonia in chest radiographs: A cross-sectional study. PLoS Med. 15, e1002683 (2018).","journal-title":"PLoS Med."},{"key":"2819_CR28","doi-asserted-by":"publisher","unstructured":"A van den Oord, Y Li, O Vinyals. Representation Learning with Contrastive Predictive Coding. Preprint at https:\/\/doi.org\/10.48550\/arXiv.1807.03748 (2019).","DOI":"10.48550\/arXiv.1807.03748"},{"key":"2819_CR29","unstructured":"Grill, J.-B. et al. Bootstrap Your Own Latent: A New Approach to Self-Supervised Learning. In Advances in Neural Information Processing Systems 33 (2020)."},{"key":"2819_CR30","doi-asserted-by":"publisher","DOI":"10.1016\/j.cmpb.2025.109138","volume":"274","author":"DD Kranz","year":"2026","unstructured":"Kranz, D. D. et al. Exploring latent diffusion models for ECG generation on the minute scale. Comput. Methods Prog. Biomed. 274, 109138 (2026).","journal-title":"Comput. Methods Prog. Biomed."},{"key":"2819_CR31","unstructured":"Chen, T., Kornblith, S., Norouzi, M. & Hinton, G. A Simple Framework for Contrastive Learning of Visual Representations. In Proc. 37th International Conference on Machine Learning 1597\u20131607 (PMLR, 2020)."},{"key":"2819_CR32","unstructured":"Loshchilov, I. & Hutter, F. Decoupled Weight Decay Regularization. In International Conference on Learning Representations (ICLR, 2019)"}],"container-title":["npj Digital Medicine"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41746-026-02819-8","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41746-026-02819-8.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41746-026-02819-8.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T04:13:44Z","timestamp":1780287224000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41746-026-02819-8"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,1]]},"references-count":32,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["2819"],"URL":"https:\/\/doi.org\/10.1038\/s41746-026-02819-8","relation":{},"ISSN":["2398-6352"],"issn-type":[{"value":"2398-6352","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,1]]},"assertion":[{"value":"16 February 2026","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"20 May 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"1 June 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"S.T. received grants\/fundings from FMRTS, FMEAE, EFFRE, Innovationsbank Berlin (IBB), DFG, Queen Mary Hospital, Orionpharma, Cytosorbents, B. Braun Stiftung, Charit\u00e9 and BD. S.T. received honoraria for lectures\/workshops from Orionpharma, Edwards LifeSciences, AOP, Cytosorbents, Chiesi, Philips outside the submitted manuscript.J.W.P. is a founder and major shareholder of Biomagnetik Park GmbH. He was not involved in the technical development, data analysis, or statistical evaluation of this study.The remaining authors declare no competing interests.","order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}},{"value":"During the preparation of this work, the authors used ChatGPT (OpenAI) only for the linguistic editing of the final manuscript. The authors have reviewed and confirmed the validity of the text and take full responsibility for the content of the publication.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Statement on Generative AI use"}}],"article-number":"412"}}