How accurately, reproducibly, and efficiently can we measure left ventricular indices using M-mode, 2-dimensional, and 3-dimensional echocardiography in children?

Background: Measurements of left ventricular (LV) size, mass, and function are the most common and important tasks for echocardiography in clinical practice and research in children with congenital and acquired heart diseases. There are little data to compare the utility of M-mode (MM), 2-dimensional (2D), and 3-dimensional (3D) echocardiographic techniques for quantification of LV indices. The objective of the study was to assess the accuracy, reproducibility, and efficiency of these echocardiographic methods for measurement of LV indices in children.

Methods: A prospective study was conducted in 20 consecutive children (mean 10.6 +/- 2.8 years, 11 male and 9 female subjects) using conventional MM, 2D, and real-time 3D echocardiography (RT3DE). A Sonos 7500 system (Philips Medical Systems, Andover, MA) was used. M-mode and 2DE measurements were made according to the American Society of echocardiography recommendations. To include the entire LV for volumetric measurement, full-volume 3D data sets were acquired from 4 electrocardiogram gated subvolumes. The 3DE measurements were made off-line manually using 4-plane and 8-plane algorithms by 4D Echo-View (TomTec Imaging Systems, Munich, Germany) and a semiautomated algorithm by QLAB (Philips Medical Systems). Magnetic resonance imaging studies were also performed to determine the LV indices by a disk summation method based on the Simpson principle.

Results: The correlation and agreement between MM, 2D, and RT3D echocardiography and magnetic resonance imaging measurements are good (r = 0.81-0.97) for the 3 methods. The correlation was superior for RT3DE compared with 2DE and MM. The correlation and agreement were similar for the three 3DE methods. The intra- and interobserver variabilities ranged from MM (4.3%-4.8% and 7.0%-8.7%), 2DE (3.3%-4.5% and 5.5%-7.3%), and 3DE (0.4%-2.3%, and 0.2%-4.8%). The total time (acquisition and analysis) used for MM measurements was the least compared with 2DE and 3DE. The total time for 3DE using the semiautomated algorithms was not significantly different compared with that for 2DE.

Conclusions: Our study showed that MM provides the most efficient assessment of LV indices but is the least accurate and reproducible technique compared with 2DE and 3DE. Three-dimensional echocardiography using both automated and manual analysis algorithm is superior to MM and 2DE for measurements of LV indices, and the automated 3DE algorithm is as efficient as 2DE. Therefore, 3DE using the automated algorithm is the method of choice for quantification of LV indices.