Meeting Banner
Abstract #1249

A Novel Biventricular Active Mesh Model for Measuring Cardiac Function and Geometry from Cine MRI

Chun G. Schiros1, Himanshu Gupta2, Steven G. Lloyd2, Louis J. Dell'Italia2, Thomas S. Denney Jr. 3

1Auburn University, Auburn, AL, United States; 2University of Alabama at Birmingham; 3Auburn University, Auburn , AL, United States

Approaches for quantifying cardiac function and geometry from cine MRI remain problematic because 1) systolic translocation of the right ventricular (RV) atrio-ventricular annulus is greater than that of the left ventricle (LV), resulting in unequal LV and RV stroke volumes and 2) surface models for measuring ventricular curvatures are usually based on coordinate systems with singularities at the ventricular apexes and assume a circular symmetry that is not applicable in the complex RV. To address these issues, a novel biventricular active mesh model is presented and validated in normal subjects, patients with mitral regurgitation and patients with pulmonary hypertension.

Keywords

able active addition agreement aligned among apex applicable approaches atrium auburn automatically available axis basal bigger binary bland boundary breath cardiac chamber coefficient color complex computation construct constructed contour contours coordinate correlation correspondences curvature curvatures deformed dell despite difficult difficulties distal distribution eccentric ejection enlarged equal errors existence facilitation field fitted fitting flow free gated generated generation geometric geometry gold greater heart hearts hypertension important improvement includes known larger long manually mapped matrix measured measuring mesh meshes model models modified moreover myocardium near novel optimized outflow overload pathologies patient patients penalizing plots potentially precision pressure problems prolate quantifying randomly reduced regurgitation remain repeated repetition representations require respectively response sample samples scales scanner secondly segmentation segmentations selected semi sets short significantly since singularity slices smaller smooth smoothed sphere spherical spheroidal steady stroke student subject subjects surface surfaces system systolic tract training triangulated typical usually utilizing validate variation ventricle ventricular vertex views volume