Estimation of right ventricular volume by modified echocardiographic subtraction method.

To evaluate the accuracy and clinical utility of right ventricular volume estimated by a modified echocardiographic subtraction method versus Krebs' original subtraction method, an experiment was performed on hearts excised from 25 animals (dogs, pigs, and cows) followed by a clinical study of 41 patients with heart disease. Right ventricular volume was measured by subtracting the left ventricular volume from that of the whole heart based on echocardiographic apical two- and four-chamber views by means of the area-length method. In the animal heart study, the coefficient of variation between the right ventricular volume estimated by the modified method and the true volume was +/- 13%. The regression equation was y = 0.94x + 4.15 (r = 0.987, p less than 0.001) and showed good correlation, whereas the right ventricular volume obtained by the original method underestimated the true volume (coefficient of variation = +/- 25%, y = 0.59x + 1.11; r = 0.976, p less than 0.001). In the clinical study, the coefficient of variation between right ventricular volume estimated by the modified echocardiographic method and RV volume estimated by radionuclide ventriculography was +/- 15%. The regression equation was y = 0.80x + 13.3 (r = 0.935, p less than 0.001). This correlation was better than that obtained by the original method (coefficient of variation = +/- 16%), where the regression equation was y = 0.60x + 2.43 (r = 0.888, p less than 0.001). Thus the accuracy of the modified subtraction method was validated, and this method showed a better correlation than the original method both experimentally and clinically.