Biomedical repositories such as the UK Biobank provide increasing access to prospectively collected cardiac imaging, however these data are unlabeled, which creates barriers to their use in supervised machine learning. We develop a weakly supervised deep learning model for classification of aortic valve malformations using up to 4,000 unlabeled cardiac MRI sequences. Instead of requiring highly curated training data, weak supervision relies on noisy heuristics defined by domain experts to programmatically generate large-scale, imperfect training labels. For aortic valve classification, models trained with imperfect labels substantially outperform a supervised model trained on hand-labeled MRIs. In an orthogonal validation experiment using health outcomes data, our model identifies individuals with a 1.8-fold increase in risk of a major adverse cardiac event. This work formalizes a deep learning baseline for aortic valve classification and outlines a general strategy for using weak supervision to train machine learning models using unlabeled medical images at scale.
Fries, Jason A., Varma, Paroma, Chen, Vincent S., et al., "Weakly supervised classification of aortic valve malformations using unlabeled cardiac MRI sequences," Nature Communications 10, no. 1 (2019), https://doi.org/10.1038/s41467-019-11012-3
@article{osti_1624170,
author = {Fries, Jason A. and Varma, Paroma and Chen, Vincent S. and Xiao, Ke and Tejeda, Heliodoro and Saha, Priyanka and Dunnmon, Jared and Chubb, Henry and Maskatia, Shiraz and Fiterau, Madalina and others},
title = {Weakly supervised classification of aortic valve malformations using unlabeled cardiac MRI sequences},
annote = {Biomedical repositories such as the UK Biobank provide increasing access to prospectively collected cardiac imaging, however these data are unlabeled, which creates barriers to their use in supervised machine learning. We develop a weakly supervised deep learning model for classification of aortic valve malformations using up to 4,000 unlabeled cardiac MRI sequences. Instead of requiring highly curated training data, weak supervision relies on noisy heuristics defined by domain experts to programmatically generate large-scale, imperfect training labels. For aortic valve classification, models trained with imperfect labels substantially outperform a supervised model trained on hand-labeled MRIs. In an orthogonal validation experiment using health outcomes data, our model identifies individuals with a 1.8-fold increase in risk of a major adverse cardiac event. This work formalizes a deep learning baseline for aortic valve classification and outlines a general strategy for using weak supervision to train machine learning models using unlabeled medical images at scale.},
doi = {10.1038/s41467-019-11012-3},
url = {https://www.osti.gov/biblio/1624170},
journal = {Nature Communications},
issn = {ISSN 2041-1723},
number = {1},
volume = {10},
place = {United States},
publisher = {Nature Publishing Group},
year = {2019},
month = {07}}
USDOE; Defense Advanced Research Projects Agency (DARPA); National Institutes of Health (NIH); National Science Foundation (NSF); US Department of the Navy, Office of Naval Research (ONR); Moore Foundation; Okawa Research Grant; American Family Insurance; Accenture; Toshiba; Intel
Proceedings of the Joint Conference of the 47th Annual Meeting of the ACL and the 4th International Joint Conference on Natural Language Processing of the AFNLP: Volume 2 - ACL-IJCNLP '09https://doi.org/10.3115/1690219.1690287
