In Journal of the American Society of Echocardiography : official publication of the American Society of Echocardiography
AIMS : Aortic stenosis (AS) is a degenerative valve condition that is under-diagnosed and undertreated. Detection of AS using limited 2D echocardiography could enable screening and improve appropriate referral and treatment of this condition. We aimed to develop methods for automated detection of AS from limited imaging datasets.
METHODS : Convolutional neural networks were trained, validated, and tested using limited 2D transthoracic echocardiogram (TTE) datasets. Networks were developed to accomplish two sequential tasks; 1) view identification and 2) study-level grade of AS. Balanced accuracy and area under the receiver operator curve (AUROC) were the performance metrics used.
RESULTS : Annotated images from 577 patients were included. Neural networks were trained on data from 338 patients (average N = 10,253 labeled images), validated on 119 patients (average N = 3,505 labeled images), and performance was assessed on a test sets of 120 patients (average N = 3,511 labeled images). Fully automated screening for AS was achieved with AUROC 0.96. Networks can identify no significant (no, mild, mild/moderate) AS from significant (moderate, or severe) AS with an AUROC = 0.86 and between early (mild or mild/moderate AS) and significant (moderate or severe) AS with an AUROC of 0.75. External validation of these networks in a cohort of 8502 outpatient TTEs showed that screening for AS can be achieved using parasternal long-axis imaging only with an AUROC of 0.91.
CONCLUSION : Fully-automated detection of AS using limited 2D datasets is achievable using modern neural networks. These methods lay the groundwork for a novel method for screening for AS.
Wessler Benjamin S, Huang Zhe, Long Gary, Pacifici Stefano, Prashar Nishant, Karmiy Samuel, Sandler Roman A, Sokol Joseph, Sokol Daniel B, Dehn Monica M, Maslon Luisa, Mai Eileen, Patel Ayan R, Hughes Michael C
2023-Jan-11
aortic stenosis, echocardiography, machine learning, screening, transthoracic echocardiography
