2.3. Application of 2D-Speckle-Tracking-Echocardiography (fetalHQ®)
The post-processing speckle-tracking, a semiautomated state-of-the-art technique, uses conventional grayscale B-mode frame-by-frame-analysis. As already described elsewhere, this pattern-recognizing method tracks the movement of the speckles representing the movements of the endocardium. They are generated by the interaction of the ultrasound with the endocardial border and show a unique spatial pattern for each region of the endomyocardial interface. Endomyocardial speckles follow tissue motion in all dimensions and their image-processing algorithm-based detection allows the quantitative analysis of velocities and deformations (strain, strain rate) of the myocardium during the cardiac cycle [14].
After measurement of global dimensions at end-diastole (ED) (Figure 1a, Video-Clip S1), for 2D-STE analysis, the raw cine-loop clips of the 4-CVs were loaded into the semiautomated border recognition program fetalHQ®. Anatomic M-mode (AMM) was used to place a line across the lateral right ventricle wall at the level of tricuspid annulus perpendicularly to the area of interest to define a single cardiac cycle by identifying end-systole (ES) and ED (Figure 1b). The clearest cardiac cycle with best endocardial visualization was selected by determining the first ED and the subsequent ED by marking the nadir on the AMM signal, and finally, the ES (first frame before opening the atrioventricular valve) by marking the peak on the AMM signal (Figure 1c). Once the most optimized cardiac cycle is selected, the systolic endocardial border of each ventricle was automatically tracked algorithm-based by the software in the previously defined end-systole frame along this cardiac cycle by manual identification of three landmarks (three-point-analysis): the septal and the lateral atrioventricular valve annulus as well as the apex, where the consideration of the moderator band required special attention. For the option of more in-depth analyses of the base (1-8), the middle (9-16) and the apex (17-24), both ventricles were divided into 24 segments automatically (cardiac mapping). It was ensured that the course of the generated line of the endocardial border from the right and left ventricles could be tracked from the beginning at the insertion of the atrioventricular valve either on the lateral or septal wall passing the apex and ending at the valve insertion on the opposite wall (Figure 1d and e). If fine tuning was necessary to ensure that the endocardium was adequately traced, end-systolic and -diastolic endocardial tracking were adjusted manually by each operator independently.
Finally, automatically calculated global and segmental parameters for fetal cardiac morphometry and function depicted graphically with superimposed AMM were obtained (Figure 1f). For a better evaluation of the quite complex tool fetalHQ®, the following parameters were selected and compared for both ventricles: For fetal cardiac morphometry (ventricular size and shape) ED and SI, for fetal cardiac function (ventricular contractility) endoGLS and FS. For simplicity, segments 1, 9, and 17 were selected for the three sections with their 24 segments (base, middle, apex).