Abstract #1900

Rapid and Accurate Quantification of Left Ventricular Ejection Fraction Using an Automated Segmentation Algorithm a Clinical Validation Study

Noel Christopher Codella¹, Matthew D. Cham¹, Richard Wong^1,2, Christopher W. Chu³, Keigo Kawaji^1,2, Kirsten Healey³, Martin R. Prince¹, Yi Wang^1,2, Jonathan W. Weinsaft^1,3

¹Radiology, Weill Cornell Medical College, New York, USA; ²Biomedical Engineering, Cornell University, Ithaca, NY, USA; ³Medicine/Division of Cardiology, Weill Cornell Medical College, New York, USA

Cardiac magnetic resonance (CMR) is an imaging standard for quantification of LV ejection fraction (EF) and volume. CMR typically relies on manual tracing (MT), which can be time consuming and operator-dependent. Automated segmentation holds potential for rapid LV quantification. In this study, we evaluated LV-METRIC performance versus MT among a broad unselected patient population; compared processing time by LV-METRIC to MT, and compared LV-METRIC and MT to an independent standard of LV flow quantification. LV-METRIC successfully yielded EF within 2 points of MT, reduced the processing time by fourteen folds, and agreed with an independent standard.