Validation of in vivo two-dimensional echocardiographic dimension measurements using myocardial mass estimates in dogs.

The accuracy of in vivo measurements of left ventricular wall thickness and chamber size by means of two-dimensional echocardiography was investigated by comparing left ventricular mass estimates obtained at end diastole in 15 closed-chest dogs with a wide range of left ventricular weights. The systolic and diastolic echocardiographic mass estimates were compared with the actual weights of the freshly excised, stripped left ventricles immediately following death. The mean +/- SD of the weighed mass of the excised ventricles was 104 +/- 25 gm. The mean predicted left ventricular mass with the use of the cylinder-ellipse geometric model and gross anatomic sections corresponding to the echocardiographic imaging planes was 106 +/- 26 gm (r = 0.97 compared to the actual weight). The mean echocardiographic estimate of left ventricular mass at end diastolic was 107 +/- 24 gm (r = 0.98 compared to the weight) and 105 +/- 26 gm at end systole (r = 0.95 compared to the weight). The correlation between the echocardiographic mass estimate at end diastole and the echo mass estimate at end systole was 0.95. Regression analysis failed to demonstrate a significant over- or underestimation of the actual weight by the calculated mass with the use of either the gross anatomic or the in vivo echocardiographic dimension measurements.