Volume measurement of the left ventricle in children with congenital heart defects: 3-dimensional echocardiography versus angiocardiography

Background: Volume measurement of the left ventricle is currently done by 2-dimensional echocardiography or angiocardiography. However, for the calculation of volumes by these methods, geometrical assumptions must be made. For a precise ventricular volumetry, independent of mathematical assumptions, imaging techniques as the 3-dimensional echocardiography (3D-echo) are required, which permit the imaging of the real ventricular shape. The aim of the study was therefore to detect, whether 3D-echocardiography is suitable for left ventricular volumetry in children with congenital heart disease and whether the 3D-echocadiographically measured volumes correlate sufficiently well with angiocardiographically measured left ventricular volumes.

Methods: 102 children with congenital heart disease were investigated. For angiocardiographic left ventricular volume measurement Simpson's rule was used. Results were corrected with Lange's correction factors. The 3D-echo data sets were registered with a rotating transthoracic transducer. Ventricular volumes were calculated after manual planimetry by summation of the volumes of the single slices.

Results: The left ventricular volume could be calculated by 3D-echo in 83% of patients of all ages. In comparison to angiocardiography, the measured volumes were 0.6 +/- 3. 3 ml (0.9 +/- 25.8%) or 7.1 +/- 28.4 ml (7.4 +/- 12.1%) smaller during systole or diastole, respectively. The correlation coefficients r(2) were 0.89 for systolic and 0.93 for diastolic measurements after logarithmic transformation. Pressure or volume overload did not influence significantly the difference between the two methods.

Conclusion: Transthoracic 3D-echocardiography with a rotating transmitter is feasible for volumetry of the left ventricle in most children. The volumes measured by 3D-echo were significantly smaller than those calculated from the angiocardiography by Simpson's rule. The measurements were not influenced by the kind of load of the ventricle. The correlation between the two methods is good.