Population and Study Protocol

A total of 260 subjects of both genders and across a wide range of ages, recruited from the employees of Centro Cardiologico Monzino (Milan, Italy) and their relatives, were prospectively studied. The study design aimed to include 15 to 20 normal subjects for each gender and age decile from 20 to 80 years. Exclusion criteria were poor echocardiographic image quality, obesity (body mass index ≥ 28 kg/m ² ), hypertension, diabetes mellitus, anemia, endocrinal dysfunction, and previous cardiovascular diseases. Moreover, 2D echocardiographic examinations were performed in all subjects to exclude any subclinical structural cardiac abnormalities. Subjects with trivial mitral regurgitation were included. The protocol was approved by the institutional review board, and informed consent was obtained from all participants.

Two-Dimensional Echocardiography

All echocardiographic exams were performed using an iE33 system (Philips Medical Systems, Andover, MA) using the S5-1 sector-array probe. Complete standard M-mode and 2D echocardiographic examinations were performed according to clinical laboratory practice. From the 4-chamber apical view, the M-mode cursor was positioned at the junction of the tricuspid valve plane with the RV free wall to evaluate tricuspid annular plane systolic excursion (TAPSE), defined as the difference in the displacement of the RV base from end-diastole to end-systole and used as an index of RV function.

Systolic artery pulmonary pressure was obtained using the noninvasive Doppler echocardiographic method from the systolic RV–right atrial gradient calculated from the peak velocity of systolic transtricuspid regurgitant flow signal using the modified Bernoulli equation and right atrial pressure derived by means of the inferior vena cava collapsibility index measured in the subcostal view.

Real-Time 3DTTE Imaging

Real-time 3DTTE imaging was performed at the end of the 2D examination using the same ultrasound unit and the X3-1 matrix-array probe, imaging with temporal resolution ranging from 26 to 37 frames/s. Two consecutive acquisitions were obtained (gathered over 7 consecutive cardiac cycles) for each of the 2 ventricles in full-volume mode from the 4-chamber apical view adapted to improve the visualization of the RV and LV cavities, respectively. The volumetric data sets were digitally stored and then transferred into a workstation for offline postprocessing analysis and 3D reconstruction.

Quantification of RV Volumes and Function

A commercially available dedicated system (EchoView, TomTec Imaging, Inc, Munich, Germany) equipped with four-dimensional RV analysis software was used to analyze the RV 3DTTE volumetric data sets ( Figure 1 A). Briefly, after manual selection of the central reference points of the tricuspid and mitral valve annuli and of the LV apex, the software visualized the RV cavity by the coronal, sagittal, and frontal cut planes automatically obtained from the 3D data set both at end-systole and at end-diastole. The first frame corresponding to the R wave of the electrocardiogram (cycles are aligned for R-R tracking) was used as end-diastole. The automatic identification of end-systole, however, needed adjustment by visual inspection of the cine loops, resulting in the manual selection of the frame with the smallest RV cavity size. After manual tracing of the endocardial borders, on each of the 3 cut planes at both end-diastole and end-systole, the software automatically detects the RV surfaces throughout the cardiac cycle. Manual correction to adjust the endocardial contours in each frame prior to quantification was performed when needed. Papillary muscles, moderator bands, and trabeculae were included in the cavity volume ( Figure 1 B). RV volumes were computed throughout the cardiac cycle, from which RV end-diastolic volume (RVEDV) and RV end-systolic volume (RVESV) ( Figure 1 C) were obtained as maximums and minimums, respectively. RV stroke volume (RVSV) and RV ejection fraction (RVEF) were then measured as the difference and as the percentage change of the volumes, respectively. The postprocessing quantification required 4 ± 2 minutes.

Example of 3D RV reconstruction. (A) Acquisition of the 7-beat full-volume data set from the 4-chamber apical view, adapted to visualize the right ventricle. (B) Manual tracing of the endocardial borders on end-diastolic and end-systolic frames on coronal, sagittal, and frontal cross-sections. (C) RV reconstruction is automatically computed along the cardiac cycle: end-diastolic (gray wireframe volume) and end-systolic (green solid volume) volumes, RVSV, and RVEF are obtained.

Quantification of LV Volumes and Function

The LV 3DTTE volumetric data sets were analyzed using commercial software (4D LV-Function 2.0, TomTec Imaging, Inc), as previously described. Briefly, end-diastolic and end-systolic volumes were semiautomatically derived after manual tracing of the endocardial borders in 4-chamber, 3-chamber, and 2-chamber views. From the obtained values, LV stroke volume (LVSV) and LV ejection fraction were calculated.

Statistical Analysis

All considered parameters were found to be normally distributed using the Kolmogorov-Smirnov test and are therefore presented as mean ± SD. To assess the relationship between RV volumes and RVSV with body surface area (BSA), Pearson’s correlation coefficient ( r ) was computed. To verify the dependence of age deciles and gender on BSA-normalized RV volumes, Spearman’s rank correlation coefficient (ρ) was used. Furthermore, the relationship between RVSV and LVSV was evaluated using linear regression (Pearson’s correlation coefficient) and Bland-Altman analyses.

To assess the reproducibility of the semiautomated quantification of RV volumes in a subset of 20 randomly chosen subjects, the main investigator reevaluated the same 3D data sets ≥2 weeks after the first analysis, blinded with respect to the results of the previous evaluation. For each computed parameter, intraobserver variability was then evaluated. The same subset was also evaluated by a different observer, blinded to the results obtained by the main investigator, to assess interobserver variability. Both intraobserver and interobserver variability were expressed as coefficients of variation (percentages), defined as the standard deviation of the difference between the two measurements, multiplied by 100 and divided by their mean value. In addition, to evaluate the bias and limits of intraobserver or interobserver agreement, Bland-Altman analysis was applied. Also, a test-retest analysis was performed using the second 3DTTE data set acquired during the same examination and the relevant coefficient of variation computed for each parameter.