RT Journal Article
SR Electronic
T1 Weakly supervised classification of aortic valve malformations using unlabeled cardiac MRI sequences
JF bioRxiv
FD Cold Spring Harbor Laboratory
SP 339630
DO 10.1101/339630
A1 Jason A. Fries
A1 Paroma Varma
A1 Vincent S. Chen
A1 Ke Xiao
A1 Heliodoro Tejeda
A1 Priyanka Saha
A1 Jared Dunnmon
A1 Henry Chubb
A1 Shiraz Maskatia
A1 Madalina Fiterau
A1 Scott Delp
A1 Euan Ashley
A1 Christopher Ré
A1 James R. Priest
YR 2019
UL http://biorxiv.org/content/early/2019/05/11/339630.abstract
AB 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 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.