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.

Significance With the increase in artificial intelligence in real-world applications, there is interest in building hybrid systems that take both human and machine predictions into account. Previous work has shown the benefits of separately combining the predictions of diverse machine classifiers or groups of people. Using a Bayesian modeling framework, we extend these results by systematically investigating the factors that influence the performance of hybrid combinations of human and machine classifiers while taking into account the unique ways human and algorithmic confidence is expressed.

No abstract

Background - The aortic valve is an important determinant of cardiovascular physiology and anatomic location of common human diseases. Methods - From a sample of 34,287 white British-ancestry participants, we estimated functional aortic valve area by planimetry from prospectively obtained cardiac MRI sequences of the aortic valve. Aortic valve area measurements were submitted to genome-wide association testing, followed by polygenic risk scoring and phenome-wide screening to identify genetic comorbidities. Results - A genome-wide association study of aortic valve area in these UK Biobank participants showed three significant associations, indexed by rs71190365 (chr13:50764607, DLEU1 , p=1.8×10 -9 ), rs35991305 (chr12:94191968, CRADD , p=3.4×10 -8 ) and chr17:45013271:C:T ( GOSR2 , p=5.6×10 -8 ). Replication on an independent set of 8,145 unrelated European-ancestry participants showed consistent effect sizes in all three loci, although rs35991305 did not meet nominal significance. We constructed a polygenic risk score for aortic valve area, which in a separate cohort of 311,728 individuals without imaging demonstrated that smaller aortic valve area is predictive of increased risk for aortic valve disease (Odds Ratio 1.14, p =2.3×10 -6 ). After excluding subjects with a medical diagnosis of aortic valve stenosis (remaining n=308,683 individuals), phenome-wide association of >10,000 traits showed multiple links between the polygenic score for aortic valve disease and key health-related comorbidities involving the cardiovascular system and autoimmune disease. Genetic correlation analysis supports a shared genetic etiology with between aortic valve area and birthweight along with other cardiovascular conditions. Conclusions - These results illustrate the use of automated phenotyping of cardiac imaging data from the general population to investigate the genetic etiology of aortic valve disease, perform clinical prediction, and uncover new clinical and genetic correlates of cardiac anatomy.

Enlargement of the aorta is an important risk factor for aortic aneurysm and dissection, a leading cause of morbidity in the developed world. While Mendelian genetics account for a portion of thoracic aortic disease, the contribution of common variation is not known. Using standard techniques in computer vision, we performed automated extraction of Ascending Aortic Diameter (AsAoD) from cardiac MRI of 36,021 individuals from the UK Biobank. A multi-ethnic genome wide association study and trans-ethnic meta-analysis identified 99 lead variants across 71 loci including genes related to cardiovascular development (HAND2, TBX20) and Mendelian forms of thoracic aortic disease (ELN, FBN1). A polygenic risk score predicted prevalent risk of thoracic aortic aneurysm within the UK Biobank (OR 1.50 per standard deviation (SD) polygenic risk score (PRS), p=6.30x10-03) which was validated across three additional biobanks including FinnGen, the Penn Medicine Biobank, and the Million Veterans Program (MVP) in individuals of European descent (OR 1.37 [1.31 - 1.43] per SD PRS), individuals of Hispanic descent (OR 1.40 [1.16 - 1.69] per SD PRS, p=5.6x10-04), and individuals of African American descent (OR 1.08 [1.00 - 1.18] per SD PRS, p=0.05). Within individuals of European descent who carried a diagnosis of thoracic aneurysm, the PRS was specifically predictive of the need for surgical intervention (OR 1.57 [1.15 - 2.15] per SD PRS, p=4.45x10-03). Using Mendelian Randomization our data highlight the primary causal role of blood pressure in reducing dilation of the thoracic aorta. Overall our findings link normal anatomic variation to extremes observed in Mendelian syndromes and provide a roadmap for the use of genetic determinants of human anatomy in both understanding cardiovascular development while simultaneously improving prediction and prevention of human disease.