Clinical data about heart failure make it clear that there is a need to develop a limited echocardiogram assisted by automated functions as a screening tool to identify people who have asymptomatic left ventricular dysfunction and who can potentially benefit from proven therapies to prevent the development of symptomatic HF. Our preliminary tests show that an abbreviated version of an echo consisting of only 6 views can provide enough information for the binary screening. Six expert echo readers achieved near-perfect performance in identifying normal vs. abnormal echoes using 6 views, as compared to the diagnosis based on the full echocardiogram. The ground truth about the geometry of the heart was provided by one expert echo reader in the form of drawings and measurements. The drawings were used as a prior in a computational model that extracts the contour of the left ventricle as the shortest path in a log-polar (complex log) representation of the ventricle. This model may represent the first step in the development of an automated function for screening echocardiogram.
Irmina Gradus-Pizlo, Kunal Agrawal, Edward Delp, Zygmunt Pizlo, "Development of screening echocardiogram for detection of asymptomatic left ventricular dysfunction" in Proc. IS&T Int’l. Symp. on Electronic Imaging: Computational Imaging XVI, 2018, pp 200-1 - 2007, https://doi.org/10.2352/ISSN.2470-1173.2018.15.COIMG-200