{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,7,3]],"date-time":"2026-07-03T16:19:29Z","timestamp":1783095569377,"version":"3.54.6"},"posted":{"date-parts":[[2026,7,1]]},"group-title":"Cardiovascular Medicine","reference-count":31,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nd\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,7,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:p>Accurate measurement of intra-cardiac blood oxygen (O2) saturation is essential for cardiovascular assessment, yet current methods require invasive catheterization. T2-based cardiac magnetic resonance imaging (CMRI) enables non-invasive O2 quantification, but deep learning automation is constrained by scarce annotated data. We propose a unified self-supervised learning (SSL) framework integrating cine CMRI and T2 oximetry CMRI to learn generalizable representations without labels. Our approach pre-trains ResNet and vision transformer encoders using contrastive learning and masked image modeling on over 48,000 cardiac images. Pre-trained encoders are fine-tuned for O2 saturation regression with uncertainty quantification to enhance clinical trustworthiness. Our SSL framework significantly outperforms traditional radiomics and supervised baselines, with SimCLR pre-trained ResNet achieving a mean absolute error of 3.70, representing over 15% improvement. These findings demonstrate SSL\u2019s potential to address annotation bottlenecks in non-invasive cardiac diagnostics.<\/jats:p>","DOI":"10.64898\/2026.06.29.26356860","type":"posted-content","created":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T19:35:27Z","timestamp":1782934527000},"source":"Crossref","is-referenced-by-count":0,"title":["Leveraging Self-Supervised Learning for Non-Invasive Intra-Cardiac Magnetic Resonance Oximetry Assessment"],"prefix":"10.64898","author":[{"given":"Jiayuan","family":"Chen","sequence":"first","affiliation":[{"name":"Department of Computer Science and Engineering, The Ohio State University, Columbus, Ohio, USA"},{"name":"Department of Biomedical Informatics, The Ohio State University, Columbus, Ohio, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Thai-Hoang","family":"Pham","sequence":"additional","affiliation":[{"name":"Department of Biomedical Informatics, The Ohio State University, Columbus, Ohio, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ping","family":"Zhang","sequence":"additional","affiliation":[{"name":"Department of Computer Science and Engineering, The Ohio State University, Columbus, Ohio, USA"},{"name":"Department of Biomedical Informatics, The Ohio State University, Columbus, Ohio, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Juliet","family":"Varghese","sequence":"additional","affiliation":[{"name":"Department of Radiology, The Ohio State University, Columbus, Ohio, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026070308301530000_2026.06.29.26356860v1.1","unstructured":"DongAo Ma , Jiaxuan Pang , Michael B Gotway , and Jianming Liang , \u201cA fully open ai foundation model applied to chest radiography,\u201d Nature, pp. 1\u201311, 2025."},{"key":"2026070308301530000_2026.06.29.26356860v1.2","article-title":"Chexagent: Towards a foundation model for chest x-ray interpretation","year":"2024","journal-title":"arXiv preprint"},{"key":"2026070308301530000_2026.06.29.26356860v1.3","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-023-06555-x"},{"key":"2026070308301530000_2026.06.29.26356860v1.4","doi-asserted-by":"crossref","unstructured":"Kai Yu , Yang Zhou , Yang Bai , Zhi Da Soh , Xinxing Xu , Rick Siow Mong Goh , Ching-Yu Cheng , and Yong Liu , \u201cUrfound: Towards universal retinal foundation models via knowledge-guided masked modeling,\u201d in International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, 2024, pp. 753\u2013762.","DOI":"10.1007\/978-3-031-72390-2_70"},{"issue":"4","key":"2026070308301530000_2026.06.29.26356860v1.5","first-page":"521","article-title":"A foundation model for enhancing magnetic resonance images and downstream segmentation, registration and diagnostic tasks","volume":"9","year":"2025","journal-title":"Nature Biomedical Engineering"},{"key":"2026070308301530000_2026.06.29.26356860v1.6","doi-asserted-by":"crossref","unstructured":"Chen Chen , Chen Qin , Huaqi Qiu , Giacomo Tarroni , Jinming Duan , Wenjia Bai , and Daniel Rueckert , \u201cDeep learning for cardiac image segmentation: a review,\u201d Frontiers in cardiovascular medicine, 2020.","DOI":"10.3389\/fcvm.2020.00025"},{"issue":"2","key":"2026070308301530000_2026.06.29.26356860v1.7","doi-asserted-by":"crossref","first-page":"111","DOI":"10.1002\/jmri.10154","article-title":"Cardiac mri: recent progress and continued challenges","volume":"16","year":"2002","journal-title":"Journal of Magnetic Resonance Imaging: An Official Journal of the International Society for Magnetic Resonance in Medicine"},{"key":"2026070308301530000_2026.06.29.26356860v1.8","unstructured":"Prabhakar Shantha Rajiah , Christopher J Fran\u00e7ois , and Tim Leiner , \u201cCardiac mri: state of the art,\u201d Radiology, 2023."},{"key":"2026070308301530000_2026.06.29.26356860v1.9","doi-asserted-by":"crossref","unstructured":"Olivier Bernard , Alain Lalande , Clement Zotti , Frederick Cervenansky , Xin Yang , Pheng-Ann Heng , Irem Cetin , Karim Lekadir , Oscar Camara , Miguel Angel Gonzalez Ballester , et al., \u201cDeep learning techniques for automatic mri cardiac multi-structures segmentation and diagnosis: is the problem solved?,\u201d IEEE transactions on medical imaging, 2018.","DOI":"10.1109\/TMI.2018.2837502"},{"key":"2026070308301530000_2026.06.29.26356860v1.10","doi-asserted-by":"crossref","unstructured":"Siying Ma , Junjie Zhou , Cailing Pu , Fuyan Wang , Qiuhui Hu , Yan Wu , Feidan Yu , Lingjie Zhang , Chengbin He , and Hongjie Hu , \u201cCmr-based t2 rv\/lv blood pool ratio: Help for risk stratification in heart failure,\u201d European Journal of Radiology, p. 112178, 2025.","DOI":"10.1016\/j.ejrad.2025.112178"},{"key":"2026070308301530000_2026.06.29.26356860v1.11","doi-asserted-by":"crossref","unstructured":"Juliet Varghese , Matthew Smyke , Yue Pan , Saurabh Rajpal , Jason Craft , Lee C Potter , Subha V Raman , Rizwan Ahmad , and Orlando P Simonetti , \u201cPatient-adaptive magnetic resonance oximetry: Comparison with invasive catheter measurement of blood oxygen saturation in patients with cardiovascular disease,\u201d Journal of Magnetic Resonance Imaging, vol. 52, 2020.","DOI":"10.1002\/jmri.27179"},{"key":"2026070308301530000_2026.06.29.26356860v1.12","unstructured":"Kaiming He , Xiangyu Zhang , Shaoqing Ren , and Jian Sun , \u201cDeep residual learning for image recognition,\u201d in CVPR, 2016."},{"key":"2026070308301530000_2026.06.29.26356860v1.13","article-title":"An image is worth 16x16 words: Transformers for image recognition at scale","year":"2020","journal-title":"arXiv preprint"},{"key":"2026070308301530000_2026.06.29.26356860v1.14","doi-asserted-by":"crossref","unstructured":"Jia Deng , Wei Dong , Richard Socher , Li-Jia Li , Kai Li , and Li Fei-Fei , \u201cImagenet: A large-scale hierarchical image database,\u201d in 2009 IEEE CVPR. IEEE, 2009, pp. 248\u2013255.","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"2026070308301530000_2026.06.29.26356860v1.15","unstructured":"Kaiming He , Ross Girshick , and Piotr Doll\u00e1r , \u201cRethinking imagenet pre-training,\u201d in Proceedings of the IEEE\/CVF international conference on computer vision, 2019, pp. 4918\u20134927."},{"key":"2026070308301530000_2026.06.29.26356860v1.16","first-page":"9912","article-title":"Unsupervised learning of visual features by contrasting cluster assignments","volume":"33","year":"2020","journal-title":"Advances in neural information processing systems"},{"key":"2026070308301530000_2026.06.29.26356860v1.17","unstructured":"Kaiming He , Xinlei Chen , Saining Xie , Yanghao Li , Piotr Doll\u00e1r , and Ross Girshick , \u201cMasked autoencoders are scalable vision learners,\u201d in Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, 2022, pp. 16000\u201316009."},{"key":"2026070308301530000_2026.06.29.26356860v1.18","doi-asserted-by":"crossref","unstructured":"P. Radau , Y. Lu , K. Connelly , G. Paul , A. Dick , and G. Wright , \u201cEvaluation framework for algorithms segmenting short axis cardiac mri.,\u201d 2009.","DOI":"10.54294\/g80ruo"},{"key":"2026070308301530000_2026.06.29.26356860v1.19","doi-asserted-by":"publisher","DOI":"10.1002\/jmri.1880010303"},{"issue":"1","key":"2026070308301530000_2026.06.29.26356860v1.20","doi-asserted-by":"crossref","first-page":"47","DOI":"10.1002\/mrm.10355","article-title":"Comparison of the dependence of blood r2 and r on oxygen saturation at 1.5 and 4.7 tesla","volume":"49","year":"2003","journal-title":"Magnetic Resonance in Medicine: An Official Journal of the International Society for Magnetic Resonance in Medicine"},{"issue":"2","key":"2026070308301530000_2026.06.29.26356860v1.21","first-page":"251","article-title":"Blood-oxygen-level dependent (bold) t2-mapping reflects invasively measured central venous oxygen saturation in cardiovascular patients","volume":"16","year":"2023","journal-title":"Cardio-vascular Imaging"},{"key":"2026070308301530000_2026.06.29.26356860v1.22","doi-asserted-by":"crossref","unstructured":"Nikita Nair , Rui Li , Gareth Matthews , Orlando Simonetti , Juliet Varghese , and Pankaj Garg , \u201cFeasibility of mr oximetry from a routine myocardial t2 mapping protocol in hfpef and hfref cohorts,\u201d Journal of Cardiovascular Magnetic Resonance, vol. 27, 2025.","DOI":"10.1016\/j.jocmr.2024.101369"},{"issue":"1","key":"2026070308301530000_2026.06.29.26356860v1.23","doi-asserted-by":"crossref","first-page":"70","DOI":"10.1186\/s12968-019-0579-7","article-title":"Free breathing three-dimensional cardiac quantitative susceptibility mapping for differential cardiac chamber blood oxygenation\u2013initial validation in patients with cardiovascular disease inclusive of direct comparison to invasive catheterization","volume":"21","year":"2019","journal-title":"Journal of Cardiovascular Magnetic Resonance"},{"issue":"3","key":"2026070308301530000_2026.06.29.26356860v1.24","doi-asserted-by":"crossref","first-page":"1545","DOI":"10.1002\/mrm.26808","article-title":"Cardiac quantitative susceptibility mapping (qsm) for heart chamber oxygenation","volume":"79","year":"2018","journal-title":"\u201dMagnetic resonance in medicine"},{"key":"2026070308301530000_2026.06.29.26356860v1.25","doi-asserted-by":"crossref","unstructured":"Joost JM Van Griethuysen , Andriy Fedorov , Chintan Parmar , Ahmed Hosny , Nicole Aucoin , Vivek Narayan , Regina GH Beets-Tan , Jean-Christophe Fillion-Robin , Steve Pieper , and Hugo JWL Aerts , \u201cComputational radiomics system to decode the radiographic phenotype,\u201d Cancer research, 2017.","DOI":"10.1158\/0008-5472.CAN-17-0339"},{"key":"2026070308301530000_2026.06.29.26356860v1.26","unstructured":"Ting Chen , Simon Kornblith , Mohammad Norouzi , and Geoffrey Hinton , \u201cA simple framework for contrastive learning of visual representations,\u201d in ICML, 2020."},{"key":"2026070308301530000_2026.06.29.26356860v1.27","doi-asserted-by":"crossref","unstructured":"Zongwei Zhou , Vatsal Sodha , Md Mahfuzur Rahman Siddiquee , Ruibin Feng , Nima Tajbakhsh , Michael B Gotway , and Jianming Liang , \u201cModels genesis: Generic autodidactic models for 3d medical image analysis,\u201d in International conference on medical image computing and computer-assisted intervention. Springer, 2019, pp. 384\u2013393.","DOI":"10.1007\/978-3-030-32251-9_42"},{"issue":"1","key":"2026070308301530000_2026.06.29.26356860v1.28","doi-asserted-by":"crossref","first-page":"74","DOI":"10.1038\/s41746-023-00811-0","article-title":"Self-supervised learning for medical image classification: a systematic review and implementation guidelines","volume":"6","year":"2023","journal-title":"NPJ Digital Medicine"},{"key":"2026070308301530000_2026.06.29.26356860v1.29","doi-asserted-by":"crossref","first-page":"e1045","DOI":"10.7717\/peerj-cs.1045","article-title":"Self-supervised learning methods and applications in medical imaging analysis: A survey","volume":"8","year":"2022","journal-title":"PeerJ Computer Science"},{"issue":"8","key":"2026070308301530000_2026.06.29.26356860v1.30","doi-asserted-by":"crossref","first-page":"872","DOI":"10.3390\/bioengineering12080872","article-title":"Cascaded self-supervision to advance cardiac mri segmentation in low-data regimes","volume":"12","year":"2025","journal-title":"Bioengineering"},{"key":"2026070308301530000_2026.06.29.26356860v1.31","doi-asserted-by":"publisher","DOI":"10.1097\/ALN.0b013e3181af59aa"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.06.29.26356860","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,3]],"date-time":"2026-07-03T15:31:30Z","timestamp":1783092690000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.64898\/2026.06.29.26356860"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,1]]},"references-count":31,"URL":"https:\/\/doi.org\/10.64898\/2026.06.29.26356860","relation":{},"subject":[],"published":{"date-parts":[[2026,7,1]]},"subtype":"preprint"}}