In vitro validation of right ventricular volume measurement by three dimensional echocardiography

Evaluation of ability of three dimensional echocardiography to accurately assess right ventricular volumes in vitro.

Silicone casts of normal human right ventricles were examined. Each was filled with three different volumes of water to yield 15 different measurements. The casts were examined in a waterbath with three dimensional echocardiography using a 7.5 MHz ultrasound probe mounted in a scan frame. It was steered by a stepper motor, which moved the probe in steps of 0.25 mm over a distance of 5.9 cm inside the frame, acquiring an image at each step. 236 parallel slices of the cast were thus obtained, forming the three dimensional dataset. The longest axis of the right ventricular volume was defined and the area of perpendicular 1 mm thick slices was outlined manually to calculate the area of each slice. This was multiplied by the slice thickness to obtain the volume of each slice; the respective volumes were added to obtain the volume of the whole cast.

The casts had a median volume of 31.1 (23) ml (range 15-100); three dimensional echocardiography gave a median volume of 29.0 (21.7) ml (15.7-91.7). Interobserver variability was 4.5% (0.4%-13.6%) and intraobserver variability 4.3% (0.2%-9.3%). Correlation between real cast volumes and volumes measured by three dimensional echocardiography was 0.99 (y = 1.08 x -0.16) with an SEE of 2.7 ml. Limits for agreement between methods ranged from -3.1 ml to 8.3 ml. In 14 of the 15 measurements, volume by three dimensional echocardiography was smaller than real volume, with the mean difference being 7.4% (2.8%-19.5%). This may be due to the thickening of surfaces of structures when imaged by ultrasonography.

Right ventricular volumes can accurately be determined by three dimensional echocardiography.