Three-dimensional echocardiography compared to two-dimensional echocardiography for measurement of left ventricular mass anatomic validation in an open chest canine model.

A three-dimensional echocardiographic system has been developed that can accurately compute left ventricular mass in vitro. This study was designed to validate the new echocardiographic system for the measurement of left ventricular mass in vivo and to compare the accuracy of three-dimensional echocardiography to the accuracy of conventional two-dimensional echocardiography for measuring left ventricular mass. Echocardiographic imaging was performed 6 h following coronary ligation in 20 open chest dogs, at which time the heart was excised and the left ventricle weighed. Three-dimensional echocardiography used multiple short axis sections and polyhedral surface reconstruction to compute myocardial volume. The two dimensional methods employed the truncated ellipsoid model and the area-length model. Myocardial volume was multiplied by 1.05 g/cc and echocardiographic mass estimates were compared to the true left ventricular weight. Three-dimensional echocardiography provided the best correlation (r = 0.96, upsilon r = 0.88 and r = 0.83 for the truncated-ellipsoid and area-length methods, respectively), and the lowest standard error of the estimate for the regression equation (+/- 5.5 g upsilon +/- 11.0 and +/- 14.6 g, respectively). Three dimensional echocardiography also had the lowest standard deviation for the echo-true mass differences (+/- 5.8 g upsilon +/- 10.7 g and +/- 14.2 g) and a lower root mean square percent error (6.8%) upsilon 12.6% and 12.7%). In this open chest canine model, three-dimensional echocardiography is more accurate than standard two-dimensional echocardiographic methods for measuring left ventricular mass.