Background: Most biplane methods for the echocardiographic calculation of left ventricular volumes assume orthogonality between paired views from the apical window. Our aim was to study the accuracy of biplane left ventricular volume calculations when either the apical two-chamber or long-axis views are combined with the four-chamber view. The left ventricular volumes calculated from three-dimensional echocardiographic data sets were used as a reference. Twenty-seven patients underwent precordial three-dimensional echocardiography using rotational acquisition of planes at 2-degree intervals, with ECG and respiratory gating. End-diastolic and end-systolic left ventricular volumes and ejection fraction on three-dimensional echocardiography were calculated by (1) Simpson's methods (3DS) at 3 mm short-axis slice thickness (reference method) and by (2) biplane ellipse from paired views using either apical four- and two-chamber views (BE-A) or apical four- and long-axis views (BE-B). Observer variabilities were studied by the standard error of the estimate % (SEE) in 19 patients for all methods.
Results: The spatial angles (mean +/- SD) between the apical two-chamber, long-axis and four-chamber views were 63.3 degrees +/- 19.7 and 99.1 degrees +/- 25.6, respectively. The mean +/- SD of end-diastolic and end-systolic left ventricular volumes (ml) and ejection fraction (%) by 3DS were 142.2 +/- 60.9, 91.8 +/- 59.6 and 39.6 +/- 17.5, while that by BE-A were 126.7 +/- 60.4, 84.0 +/- 57.9 and 39 +/- 17 and by BE-B were 134.3 +/- 62.4, 88.6 +/- 59.7 and 39.1 +/- 16.7, respectively. BE-B intra-observer (8.4, 6.7 and 3.5) and inter-observer (9.8, 11.5 and 5.4) SEE for end-diastolic and end-systolic left ventricular volumes (ml) and ejection fraction (%), respectively, were smaller than that for BE-A (10.8, 8.8 and 4.1 and 11.4, 14.7 and 6.1, respectively). There was excellent correlation between 3DS and BE-A (r = 0.99, 0.98 and 0.98) and BE-B (0.98, 0.98 and 0.98) for calculating end-diastolic and end-systolic left ventricular volume and ejection fractions, respectively. There were no significant differences between BE-A and BE-B with 3DS for end-diastolic and end-systolic left ventricular volume and ejection fraction calculations (P = 0.2, 0.3 and 0.4 and P = 0.5, 0.5 and 0.4, respectively). There were closer limits of agreement (mean +/- 2 SD) between 3DS and BE-B 7.9 +/- 18.8, 3.2 +/- 14.2 and 0.8 +/- 5.8 than that between 3DS and BE-A 15.5 +/- 19.6, 7.8 +/- 16.2 and 1.1 +/- 7.4 for calculating end-diastolic and end-systolic left ventricular volume and ejection fractions, respectively.
Conclusion: Both apical two-chamber and apical long-axis views are not orthogonal to the apical four-chamber view. Observer variabilities of BE-B were smaller than that for BE-A. BE-A and BE-B have excellent correlation and non-significant differences with 3DS for left ventricular volume and ejection fraction calculations. There were closer limits of agreement between BE-B with 3DS for left ventricular volume and ejection fraction calculations than that between BE-A and 3DS. Therefore, we recommend the use of the apical long-axis rather than the two-chamber view in combination with the four-chamber view for accurate biplane left ventricular volume and ejection fraction calculations.