Remarkable advances have taken place in three-dimensional echocardiography (3DE) in congenital heart disease over the past two decades. The technique has progressed from a cumbersome modality involving, for example, slow rotation of ultrasound probes to reconstitute a virtual image, to a live technique that can produce images in real time to assess cardiac anatomy, function and blood flow. Despite these major advances, there remains considerable scope for further research and innovation to enhance the technique still further. The areas for advancement include not only technical issues, but also agreement on some core standards of practice, analogous to those that have already been widely adopted for cross-sectional echocardiography, and have recently been published for acquired disease in adult practice .
One of the major attractions of 3DE is the ability to project images of the true living anatomy from unique projections that can be achieved during post processing. These images illustrate the “full picture” in contrast to the specific sonographic cuts of cross-sectional echocardiography, thereby facilitating visualization of the anatomy of the heart, with the aim of assisting either the surgeon or the interventional cardiologist. Challenges that remain include the development of a consistent means of image orientation for the more common en face views of valves and septal structures, so that a “common language” is adopted between the echocardiologist, surgeon and interventionist. One option would be the adoption of an anatomical approach to image display , which would be consistent with imaging modalities such as computed tomography (CT) and magnetic resonance imaging (MRI), and would be logical with respect to the relative position of cardiac structures in free space. This question is currently under consideration by the European Association of Echocardiography and the American Society of Echocardiography, with the aim of developing a consensus-based approach.
The understanding of atrioventricular valve function by 3DE has advanced beyond qualitative visualization of the atrioventricular valve and chordal apparatus, to a much fuller understanding of valve function by quantitative assessment of annular size and shape, valvar tethering and the crucial interplay of the valve leaflets, and the geometry of the chordal support apparatus . Advances in this area would be facilitated by the development of commercially available software that does not make assumptions about valve anatomy or morphology. Such software has been extensively developed for acquired abnormalities of normal valves, but is lacking for the congenital population; as a result, many reports have relied upon “research” software that cannot be used readily in more widespread clinical practice. Similarly, software to quantitate valvar regurgitation – for example, by effective regurgitant orifice area–needs to be adapted to the congenitally abnormal heart, where the mechanism of regurgitation and regurgitant area may differ significantly from acquired valve lesions. Such tools will be essential to advance understanding of areas of major clinical importance, such as mechanisms of atrioventricular valve regurgitation in the context of single ventricle circulation or late failure of previously repaired valves.
Assessment of ventricular volumes and function in the congenitally abnormal heart has been a central goal of 3DE since inception; this is because 3DE makes few assumptions about ventricular geometry compared with M-mode or cross-sectional echocardiographic techniques. This feature is particularly pertinent for the morphologic right ventricle and for single ventricle physiology, where the shape of ventricles may be highly unusual and where the usual anatomical landmarks may be absent. Application of the technique has been limited by the availability of suitable software and practical difficulties, such as incorporation of the entire ventricle into a single 3DE volume. Most studies have compared the echocardiographic technique with cardiac MRI, and although the techniques correlate reasonably well, there is a significant trend for the echocardiographic technique to produce lower ventricular volumes than MRI, and for the limits of agreement to be too wide to use the techniques interchangeably . Novel solutions have been proposed to overcome such limitations, including “knowledge-based reconstruction”, where a tracked probe and multiple cross-sectional images are used to reconstruct the three-dimensional (3D) ventricle by reference to a library of datasets of the same lesion . Other research strategies have included fusing multiple 3D datasets together to cover the entire ventricle, but this remains a research rather than a clinical technique. The challenge over the coming years will be to develop user-friendly systems that overcome these current limitations, so that there is increased confidence in 3DE volume estimation in the congenitally abnormal heart; this has the potential to reduce the requirement for other techniques, such as MRI, for ongoing monitoring.
To date, most emphasis has been placed on estimation of ventricular volumes and ejection fraction, but in recent years, a much greater understanding has emerged of intracavity flow, in particular vortex formation, which impacts on ventricular efficiency and energy loss ( Fig. 1 ). These are extremely important considerations in the congenital heart disease population, particularly in the group with single ventricle physiology, where energy conservation within the blood pool is central to optimization of cardiac output. 3DE can be used to evaluate intracavity flow in a research setting, but is not sufficiently developed to be a current clinical technique, although this remains a future aim . It is likely that advances in 3DE assessment of cardiac function will not be restricted to either endocardial border detection or intracavity flow. Two-dimensional (2D) strain (speckle tracking) represents a significant advance, given the ease of acquisition (particularly important in younger patients) and rapid post processing of data. Myocardial fibre orientation is well recognised as being abnormal in many forms of congenital heart disease. 3DE wall tracking is now technically feasible and avoids the potential problems of 2D techniques, such as through plane motion of the myocardium. Recently, age-related normal values for 3D strain, myocardial rotation and torsion have been published in children and adolescents , which should assist with the adoption of the technique in clinical practice. Ongoing challenges include the differences in strain values between vendors and between 2D and 3D strain techniques.
