OUR EXPERIENCE

Zosmer, Campbell et al have demonstrated, under favorable conditions, the validity of volumetric data in the examination of some heart scans. These authors were able to obtain data on the apical and lateral volumes of the fetal heart of 54 patients, and have subsequently re-examined the images produced. These patients’ gestational ages were between 17 and 37 weeks; the most suitable scans were chosen for diagnosis. The main scans sought were the following:

• 4 chambers;
• Left long-axis;
• Aortic crest;
• Short-axis;
• Ductal arch;

Surprisingly, the movements of the cardiac muscle and valve cause slight distortion and thus provide little indication on the quality of the ultrasonographic volume and the plane scans of the apical volumes. On the other hand the quality of the lateral cardiac volumes obtained is compromised by the cardiac movements. The most suitable period for a multiplanar (MP) echocardiographic examination seems to be between the 22nd and 27th gestation week. Volumetric multiplanar ultrasound allows the reproduction of an ultrasound image of the fetal corporeal volume which can be easily examined with tomographic images, allowing the simultaneous comparison of the voxel point on multiplanar scans. These are particularly useful for structured anatomy examinations of regions with a defined spatial organization and measuring more than 2 or 3 mm (the limit being the ultrasound resolution) like fetal brain and heart. The use of these instruments is of great advantage when studying the heart, provided that the volume is significant. The possibility to compare one point on two or three orthogonal scans allows a more accurate examination of the cardiac connection, and more significant images, from a medic-legal point of view, than the B-Mode examination.
The advantages of a fetal echocardiacography using volumetric examination can be summarized as follows:
1. The possibility to examine and compare the position of the same point (voxel point) on 3 orthogonal planes. This is particularly useful to study fetal cardiac connections and to produce more exhaustive and precise comparative images that are more useful for medical legal purposes;
2. Volumes may be stored on 2 Gb Jaz discs, or on IBM-compatible PC by transferring the volume by a DICOM system, so as to carry out the examination at the end of the pregnancy or even later, if necessary;
3. Reduced examination time;
4. The data on the cardiac volume may be submitted to a more experienced examiner or to other centres (by storing the data on Syquest/Jaz removable disk or transmitting the data stored on the PC via telephone connection);
5. The data on fetal cardiac volume at a given gestational age may be stored for future retrieval. As a matter of fact, since some prenatal fetal heart pathologies can evolve, it may be useful to be able to evaluate a posteriori prenatal conditions for any possible medico-legal controversies;
6. Acquisition of material for didactic purposes (training of operators who specialise in fetal echocardiography.

Our experience with 125 pregnancies confirms and partly expands Zosmer’s results. The scans we obtained and examined (see below) by a single volume are as follows:

• 4 chambers;
• Long left axis with voxel point, for multiplanar comparison, on the corresponding oblique short axis;
• Crossing of the long right axis with voxel point, for multiplanar comparison, on the corresponding oblique short axis;
• Crossing with the long right axis
• Right short axis;
• Ventricular short axis;
• Aortic arch with neck vessel;
• Pulmonary arch continuing into the ductus arteriosus;
• Pulmonary veins in the left atrium: The use of reformatted scans allow us to visualise a coronal section of the 4 pulmonary veins in the left atrium, and to compare the position of the voxel point on the corresponding transverse scan of the 4 chambers;
• Superior and inferior vena cava in right atrium, allowing us to compare the voxel point on the corresponding transverse scan of the 4 chambers.
The undergoing study still requires a statistical validation; nevertheless, its provisional results show that in 72% of cases it enabled a complete or nearly complete morpho-volumetric cardiac examination of fetal heart based on a single volume acquisition.

When the ultrasonographic volume is significant, it is possible to perform an exhaustive morpho-volumetric ultrasonography that traces the phases of a B-Mode Real-Time examination:

• Examination of the veno-atrial connection;
• Examination of the ventricular arterial connection;
• Examination of the atrio-ventricular connection and the atrio-ventricular valves.

The diagnostic value of the images of the cuspid valves, obtained with the volume examination, is still debated. As a matter of fact, Zosmer underlines that during a rapid volumetric acquisition (approx. 3 seconds) about 4 to 5 cardiac cycles may occur, which can invalidate the quality of the valvular images obtained. Actually, considering that the minimum time for a volumetric image corresponds to 3 seconds (about 4 to 5 cardiac cycles), the time needed to obtain the images of the single valve corresponds to less than 0.20 seconds, which is equal to less than half one cardiac cycle. Therefore the image of the valve can be influenced by negligible artifacts.
Actually pseudo-quadrimensional ultrasound has recently been introduced in commercially available V-Mode machines (see MOVIE below): unfortunately the acquisition system of multiple scans, in every single “slide” of the acquired volume, are not synchronized with fetal cardiac cycle: therefore the movies presents artifacts and are not actually useful neither for fetal multiplanar echocardiography nor for live 3D rendering of cardiac structures physiologically in movement, while in post-natal life, the synchronization of the acquired images with the cardiac cycle allows a good 3D rendering of post-natal heart.

In any case the evaluation of valvular movements through MP ultrasound is not actually advisable. The evaluation of the veno-atrial and ventricular-arterial connections can, however, in some conditions, take advantage of a volumetric examination of the fetal heart.
In our recent study we have examined the pulmonary veins connections to left atrium: it was possible to document the 4 pulmonary veins in LA in 46% of cases (Varvarigos E, Iaccarino M et al.: Multiplanar ultrasound in the identification of four pulmonary veins in left atrium. Abstract - ISUOG World Congress. New York, 2002).
The current limitations of the volumetric examination in fetal echocardiography consist of the acquisition mechanism: It is therefore necessary to perform a 'real-time volume' assessment at the end of the study, or to use any other method that allows the acquisition of a single volume in less than 0.1 second and to obtain a sequence of volumetric frames at the end of a study of the cardiac valves on both multiplanar volumetric examination and in 3D.

WHAT ARE WE EVALUATING?
Our team is studying cardiac ventricular ejection: The pulsatory ejection is the product of TPI (the integral of the Doppler spectrum curve) multiplied by the area defined by the aortic valve and pulmonary annulus, resulting by the reformatted plane sections showing the valvular circumference. The measurement of the resulting area is a reliable datum as it is taken from the measurement of the diameter or radius, which are calculated by regression to the circumference and not by direct measurement. In addition, we are devising a method to measure the residual cardiac volume (calculated from the segmentation of the cardiac chamber taken on 3 scans) and the holosystolic volume ejected from the heart (calculated as reported above).