The study group consisted of 34 healthy subjects (mean age: 36.1±11.2 years, 15 men) in sinus rhythm. None had known disease or any factor which could theoretically affect the results. Data on these subjects were taken from the MAGYAR-Healthy study (Motion Analysis of the heart and Great vessels bY three-dimensionAl speckle tRacking echocardiography in Healthy subjects) with the aim of clarifying the diagnostic and prognostic significance of 3DSTE-derived volume, strain, rotation and dyssynchrony parameters in healthy volunteers. Informed consent was obtained from all subjects. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the local human research ethics committee.

All 3DSTE studies were performed with a Toshiba Artida echocardiograph (Toshiba Medical Systems, Tokyo, Japan) using a 1–4 MHz PST-25SX matrix phased-array transducer.2 Acquisition of a full-volume 3D dataset required smaller wedge-shaped subvolumes from six consecutive cardiac cycles obtained during a single breathhold, which were then combined to provide the larger pyramidal 3D volume. LA quantification was performed using 3D Wall Motion Tracking software, version 2.7 (Toshiba Medical Systems, Tokyo, Japan). The 3D datasets were displayed in five different cross-sections comprising apical 2- (AP2CH) and 4-chamber (AP4CH) views and three standard short-axis views at different LA levels from the base to the apex. The orientation of the long axis of the AP2CH and AP4CH views was determined by positioning the main axis line to pass near the center of the LA cavity. The three short-axis views were defined by positioning the lines in AP2CH and AP4CH views at each level perpendicular to the LA long axis. The LA cavity was traced on the endocardium in AP2CH and AP4CH views starting at the edge of the septal mitral annulus (at the origin of the anterior mitral leaflet), then markers were placed in a counterclockwise rotation around the LA to the lateral mitral valve ring (to the origin of the posterior mitral leaflet). The LA epicardial border was manually adjusted. After detection of boundaries at the end-diastolic reference frame, wall motion tracking was then automatically performed through the entire cardiac cycle. A 3D cast together with volumetric and functional parameters of the LA were then generated (Figure 1).

Figure 1.

The three-dimensional echocardiographic dataset is displayed in apical 4-chamber (A) and 2-chamber views (B) and three short-axis views in the basal (C3), mid-atrial (C5), and superior (C7) regions, respectively. Three-dimensional LA cast (D), LA volumetric data, ejection fraction (EF) (E) and global peak longitudinal strain (white arrow) and global longitudinal strain at atrial contraction (dashed arrow) (F) are also presented. LA: left atrium; LV: left ventricle; RA: right atrium; RV: right ventricle.

(0.54MB).

From the 3D model of the LA, maximum LA volume (at end-systole, largest LA volume before mitral valve opening [Vmax]), minimum LA volume (at end-diastole, smallest LA volume before mitral valve closure [Vmin]) and LA volume before atrial contraction (at time of P wave on ECG [VpreA]) were calculated (Figure 1).5–9 To characterize the reservoir, conduit and active contraction phases of LA function, stroke volumes (SV) and emptying fractions (EF) were measured from the three volumes using the following equations:

Reservoir function:

• -

  Total atrial stroke volume (TASV): Vmax−Vmin

• -

  Total atrial emptying fraction (TAEF): TASV/Vmax×100

Conduit function:

• -

  Passive atrial stroke volume (PASV): Vmax−VpreA

• -

  Passive atrial emptying fraction (PAEF): PASV/Vmax×100

Active contraction:

• -

  Active atrial stroke volume (AASV): VpreA−Vmin

• -

  Active atrial emptying fraction (AAEF): AASV/VpreA×100.

The main advantage of 3DSTE is that from the same 3D model of the LA, several functional parameters including strains can be easily measured.7–11 On the basis of one-directional radial, longitudinal and circumferential strains, area strain (ratio of endocardial area change during the cardiac cycle) and 3D strain (strain in the wall thickening direction, combination of one-directional strains) can also be calculated. Global peak strains (characterizing LA reservoir function) and global strains at atrial contraction (characterizing LA active contraction function) were measured for each subject (Figure 1).

There is a third way to analyze LA function, by calculating LAEF. According to Newton's second law of motion, the force generated by the LA in its active contraction phase can be calculated using the following equation: LAEF=0.5×1.06×(MAD or MAA)×V2, where 0.5 is a coefficient, 1.06 g/cm3 is the blood density, MAD is the mitral annulus diameter, MAA is the mitral annulus area, and V is the peak A wave velocity.13 From the same 3D echocardiographic dataset, the mitral annulus (MA) can be obtained by optimizing cross-sectional planes on the endpoints of the MA in AP4CH and AP2CH views12 (Figure 2). MAD is then defined as the perpendicular line drawn from the top of the MA curvature to the middle of the straight MA border, while MAA can also be measured using planimetry. For measurement of LAEF, diastolic MAD and MAA data were used.

Figure 2.

From the three-dimensional echocardiographic dataset, the mitral annulus (MA) can be obtained by optimizing cross-sectional planes in apical 4-chamber (A) and 2-chamber (B) views, demonstrating an optimal MA image on cross-sectional view (C7). Using Doppler-derived mitral inflow peak A wave velocity, the left atrial ejection force (LAEF) can be calculated. E and A: Doppler-derived mitral inflow velocities; LA: left atrium; LV: left ventricle; MA: mitral annulus; RA: right atrium; RV: right ventricle.

(0.48MB).

Continuous variables are expressed as mean values ± standard deviation, while categorical data are summarized as percentages. A p value of <0.05 was considered to indicate statistical significance. Pearson's coefficient was used for correlations. Recently, intra- and interobserver agreement for LA volumes and functional properties were assessed in papers based on the MAGYAR-Healthy and MAGYAR-Path studies.6,8 However, 3DSTE-derived MAD and MAA were not assessed. Intra- and interobserver agreements were studied according to Bland and Altman's method.14 All calculations were performed with commercially available software (MedCalc, Mariakerke, Belgium).

Baseline clinical and echocardiographic data for the study population are presented in Table 1. All two-dimensional (2D) echocardiographic and 3DSTE-derived data were in normal ranges in this healthy population.

Volume-based and strain parameters derived from 3DSTE characterizing all phases of LA function together with LAEF are presented in Table 2.

Both MAD- and MAA-based LAEF showed correlations with global 3D strain at atrial contraction, and MAD-based LAEF correlated with AAEF (Table 3). Although LAEF is a characteristic of LA booster pump function, correlations could be demonstrated between LAEF and volume-based and strain characteristics of LA reservoir function, as well global peak strains, TASV and TAEF. No correlation could be demonstrated between LAEF and parameters characterizing LA conduit function (PASV, PAEF).

Reproducibility measurements were performed in 17 healthy controls. The mean ± standard deviation differences in values obtained by two observers for the measurements of 3DSTE-derived diastolic and systolic MAD and diastolic and systolic MAA were −0.02±0.43 cm, 0.02±0.43 cm, −0.06±1.49 cm2 and 0.07±1.02 cm2, respectively. Correlation coefficients between measurements of two observers were 0.77, 0.79, 0.89 and 0.89 (p=0.0003, 0.0002, <0.0001 and <0.0001), respectively (interobserver agreement) (Figures 2–5). The mean ± standard deviation differences in values obtained in two measurements by the same observer for 3DSTE-derived diastolic and systolic MAD and diastolic and systolic MAA were 0.03±0.38 cm, −0.01±0.34 cm, 0.05±0.87 cm2, and 0.04±0.97 cm2, respectively. Correlation coefficients between these independent measurements by the same observer were 0.78, 0.83, 0.96 and 0.90 (p=0.0002, <0.0001, <0.0001 and <0.0001), respectively (intraobserver agreement) (Figures 3–6).

Figure 3.

Interobserver (upper graphs) and intraobserver (lower graphs) agreements and correlations for measuring end-diastolic mitral annulus diameter by three-dimensional speckle tracking echocardiography are presented. 3DSTE: three-dimensional speckle tracking echocardiography; MAD: 3DSTE-derived mitral annulus diameter; obs: observer.

(0.55MB).

Figure 4.

Interobserver (upper graphs) and intraobserver (lower graphs) agreements and correlations for measuring end-systolic mitral annulus diameter by three-dimensional speckle tracking echocardiography are presented. 3DSTE: three-dimensional speckle tracking echocardiography; MAD: 3DSTE-derived mitral annulus diameter; obs: observer.

(0.48MB).

Figure 5.

Interobserver (upper graphs) and intraobserver (lower graphs) agreements and correlations for measuring end-diastolic mitral annulus area by three-dimensional speckle tracking echocardiography. 3DSTE: three-dimensional speckle tracking echocardiography; MAA: 3DSTE-derived mitral annulus area; obs: observer.

(0.51MB).

Figure 6.

Interobserver (upper graphs) and intraobserver (lower graphs) agreements and correlations for measuring end-systolic mitral annulus area by three-dimensional speckle tracking echocardiography. 3DSTE: three-dimensional speckle tracking echocardiography; MAA: 3DSTE-derived mitral annulus area; obs: observer.

(0.55MB).