Evaluation of Atrial and Ventricular Septal Defects with Real-Time Three-Dimensional Echocardiography: Current Status and Literature Review
- First Online:
- Cite this article as:
- Roberson, D.A. & Cui, V.W. Curr Cardiovasc Imaging Rep (2011) 4: 349. doi:10.1007/s12410-011-9102-8
- 98 Views
Within this report we present a state of the art assessment of the status of 3D echocardiographic imaging (3D echo) of atrial septal defect (ASD) and ventricular septal defect (VSD). After a brief literature review of past studies, we delve into more recent developments in greater detail. For ASD we focus on new developments and reports regarding real-time transesophageal echocardiography. For VSD there have been no recent reports in the last 2 years dealing with new advances in 3D echo of VSD by either transthoracic or transesophageal approaches. Therefore, we present our experience in 3D echo of VSD by both methods. The literature is far from replete on this subject; therefore, a significant portion of this report is based on the authors’ personal experience using 3D echo in over 200 cases with ASD, VSD, or both lesions in children and adults. A collection of representative reference 3D echoes of ASD and VSD are presented to serve as a review of the anatomic variations which the echocardiographer and sonographer are likely to encounter.
Keywords3-dimensional echocardiography3-dimensional transesophageal echocardiographyAtrial septal defectVentricular septal defectCongenital heart diseaseEchocardiography
The initial report of 3D echocardiographic imaging (3D echo) of ventricular septal defect (VSD) which relied on post-acquisition 3D reconstruction of 2D transthoracic images appeared in 1994 . From then until now, the value of 3D echo in general for the analysis of congenital heart disease  and in adult cardiology  has been a subject of ongoing investigation. The 3D echo technology has advanced in a somewhat predictable albeit fairly slow fashion, through the development of 3D reconstruction of 2D transesophageal images, followed by actual real-time transthoracic 3D imaging with a matrix array dedicated 3D transducer, and most recently real-time 3D TEE with a matrix array transesophageal transducer.
Due in part to the widespread success, familiarity, ease of application, and well established protocols for training, image acquisition, and analysis using 2D echocardiography, the incorporation of 3D echocardiography into routine assessment of congenital heart disease has been slow. The relatively slow to develop improvements in 3D echocardiography hardware and software have further hindered the advancement of 3D echo into the echo imaging mainstream for congenital heart disease. Recent technical developments in hardware and software continue to improve the ease of application, image quality, image processing, and quantitation capabilities, thereby alleviating some of the barriers to the more widespread application of 3D echocardiography. In the realm of adult cardiology, 3D echo has entered the mainstream as an important modality used to quantify ventricular volume and mass, synchrony, and detailed mitral valve analysis. In the congenital heart disease realm, with the possible exception of 3D transesophageal echocardiography (TEE) of atrial septal defect, the use of 3D echo appears to have a much less widespread acceptance and application, as reflected by a paucity of published literature from only a few centers.
The purpose of this report is to review the 3D echo literature regarding atrial septal defect (ASD) and VSD, describe the current state of the art 3D echo technology, and provide a collection of reference images for ASD and VSD. We focus exclusively on the anatomic analysis. Assessment of flow volume in these anomalies is not yet well developed and therefore left for subsequent reports.
Three-Dimensional Echocardiography Modes and Methods
Transthoracic 3D echo can be applied with no apparent lower bodyweight limitation. However, transesophageal 3D echo is limited to those patients >20 kg bodyweight. The currently available 3D echo modes of imaging are live 3D, 3D zoom, 3D full volume, X-plane, and 3D color Doppler flow analysis. All are available for transthoracic and transesophageal approaches. Each has its own strengths and limitations. They are described in more detail as follows.
Live 3D mode is real-time live imaging. It provides a wedge-shaped 3D sector 15° to 20° thick. The sector volume expands in height and width at greater imaging depth. Real-time continuous steering analogous to 2D echo transducer manipulation is required in live 3D mode. In order to gain maximum benefit of the 3D capabilities, the structure of interest is positioned in the far field of the echo sample volume where the 3D echo volume is largest. The volume rate is fairly rapid, occasionally achieving up to 60 volumes per second. Because there is no stitch artifact, this mode is most useful in smaller and younger patients in whom the heart rate is fast and respiratory gating or breath holding more difficult. There are two available settings consisting of slightly larger sample volume with medium resolution versus a slightly smaller sample volume with high-resolution setting. There is only a minimal difference between these two settings in clinical practice. The anterior to posterior and the left to right position of the sector can be adjusted by the elevation and lateral steering controls, respectively. Provided that the entire structure of interest can be included within the somewhat small image volumetric wedge, this is our preferred 3D TEE imaging mode. The live 3D acquisitions are often quite useful when obtained from the deep transgastric position which places the ASD or other structure of interest in the far field where the 3D echo sector is largest.
3D zoom mode is live real-time 3D imaging with user-defined 3D echo sample volume size and position. The sonographer adjusts sample volume position, depth, width, and height in order to provide a real-time pre-cropped image which can be further rotated or cropped as needed throughout all planes. This mode is quite useful to demonstrate real-time en face views of the atrial septum and ventricular septum, quantitative dynamic analysis, and determination of final closure device configuration, especially via the transesophageal route during interventional catheterization procedures. It is limited by a slow volume rate of only 5 to 18 volumes per second.
3D full-volume mode consists of rapid reconstruction, not actual real-time acquisition or viewing. A large wide-angled pyramidal volume of 3D echo is rapidly reconstructed from 4 to 7 live 3D contiguous sample volumes which are stitched together sequentially beat by beat over 4 to 7 cardiac cycles. It must be stored and cropped according to the particular region of interest. It takes the experienced 3D echocardiographer approximately 1 minute to perform most of the basic cropping functions, but can take several minutes for more complex cropping requirements. The volume rate of full-volume 3D is rapid, but the beat-by-beat EKG-triggered reconstruction is prone to stitch artifact from either rapid heart rate or respiratory motion. The latter two problems are significant in children because the stitch artifact can be severe at faster heart rates and moderate respiratory rates. It also requires post-acquisition cropping, which results in a short delay between image acquisition and the useful display of the 3D image. Therefore, unlike the live and zoom modes, it is not actual real-time echo and requires cropping of a stored digital image or cine loop, rather than on an actual live beating heart image.
X-plane 3D mode displays simultaneous orthogonal 2D or color Doppler flow map images in real time in a side-by-side format analogous to biplane angiography. The planes of dissection of the reference image and angle between the two simultaneous biplane images are adjustable and defined by the sonographer. X-plane imaging is useful to define spacial relationships and to be certain that measurements of a structure such as a valve annulus are being performed in the correct plane.
3D color Doppler flow analysis is a rapid reconstruction 3D color Doppler flow analysis modality. In similar fashion to full-volume 3D echo, images are rapidly reconstructed over 4 to 7 heart beats resulting in a sample volume approximately half the size of the standard 3D full-volume acquisition but with color Doppler flow analysis superimposed on the anatomic data. Frame rates range from approximately 20 to 45 per second and a broad array of adjustments are available.
All modes can be further cropped in three orthogonal planes or using a single or multiple infinitely adjustable planes. Multiple adjustable parameters include gain, contrast, smoothing, image algorithms, colorization adjustments, color depth shading adjustments, rotation of images throughout 180° in orthogonal planes, and more. Image manipulation, cropping, and quantification can be performed on the ultrasound platform or on a digital storage and review station. Commercially available quantitation programs provide for rapid measurement of volume, mass, length, area, synchrony, and most recently strain analysis.
When viewing the final still image it is not always apparent which 3D echo mode was used for the image acquisition. In live video images, the live 3D mode is noticeable by its small sector size, rapid volume rate, and absence of stitch artifact; the 3D zoom mode is noticeable by its slow volume rate and smooth images; and 3D full-volume mode is noticeable by its large sample volume and stitch artifact if present.
En face views are often the most useful 3D echo views of septal defects as they demonstrate the type, position, orientation, size, dynamic changes, number of orifices, device size, rim status, surrounding structures, guide interventions, and can be used to assess residual defects for both ASD and VSD. En face views can be obtained using live 3D, 3D zoom, and 3D full-volume modes. Multiple examples of these useful and intuitive views are demonstrated in the figures of this report.
Subcostal views, also called subxiphoid views, are essential to provide a complete surface echocardiogram anatomic assessment of congenital heart disease in segmental fashion  according to ASE guidelines. The transgastric views acquired during transesophageal echo are analogous to these surface echo subcostal views and are of equal import and utility for analyzing congenital heart defects.
Atrial Septal Defect
Types of ASD
3D Echo of ASD Historical Perspective
Marx et al.  first reported 3D images of secundum ASD which were reconstructed from transthoracic 2D images. The images were useful despite somewhat limited resolution and the labor-intensive reconstruction. Magni et al.  validated the accuracy of the 3D reconstruction method to measure ASD major and minor axes, size, and location in 10 explanted porcine cadaver hearts with surgically created ASD. Acar et al.  subsequently reported that the en face 3D reconstructed images of secundum ASD accurately demonstrated the shape, size, and rim features. Cao et al.  found that visualization of the number of ASD orifices, shape, and surrounding structures was enhanced using reconstructed 3D images derived from 2D TEE. In 2006, Van den Bosch et al.  first reported the use of real-time matrix array 3D transthoracic transducer which provided the currently available rapid reconstruction, not quite real-time, 3D full-volume mode acquisition images of ASD. This new technology provided better accuracy, improved image quality, and very short processing time compared to prior technology. Important information and enhanced understanding of secundum ASD anatomy and physiology were provided.
Reports of the use of real-time 3D transesophageal echo (RT3D TEE) appeared in 2009 through 2011. The progress of 3D TEE technology and its methods of application are beautifully demonstrated in these reports.
Taniguchi et al. [10•] reported results from a series of 48 patients undergoing ASD device closure. They found that RT3D TEE provided optimal images of ASD anatomy, surrounding rims, and ASD anatomic quantitation in 96% of cases. RT3D TEE was quite useful to guide device manipulation and observer variability was low.
Although not dealing precisely with ASD, Faletra et al. [11••] describe a detailed step-by-step approach for imaging of right atrial anatomy, including the atrial septum, as it applies to intracardiac electrophysiologic procedures. This report includes detailed recommendations for performing image acquisition, cropping, and instrument settings in a structure-oriented approach. Many exquisite RT3D TEE images and matching pathology specimens are included and lend credence to the contention that 3D echo can provide virtual anatomic specimen imaging.
A fine collection of RT3D TEE images of the normal atrial septum, patent foramen ovale, and representative cases of various types of atrial septal defect was presented by Pushparajah et al. [12••]. This report also demonstrates the value of using a consistent anatomically correct method of image orientation and cropping and the great utility of the en face view.
Saric et al. [13••] demonstrated the value of a standardized approach to RT3D TEE image acquisition and orientation of ASD morphology. Using the curious but effective mnemonic acronyms of TUPLE and TUPLE plus ROZL, they describe and illustrate in detail the methods to properly orient images which are acquired at 0° versus 90° transducer array orientation. Rather than wrongly recapitulate the recommended technique, I direct the reader to view this fine report directly. Its results were drawn from 23 cases of various types of ASD.
Roberson et al. [14••] reported results from a series of 65 patients including all types of ASD. RT3D TEE successfully demonstrated ASD type, location, shape, orientation, and dynamic changes in dimensions with high accuracy and low observer variability. Strengths and limitations of the 3D echo modes and recommended RT3D TEE views were also discussed.
Goals of 3D Echo of ASD
The goal of 3D echo of ASD is to determine the type, position, orientation, size, dynamic changes, number of orifices, device size, rim status, guide interventions, and assess residual defects.
3D TEE Imaging Protocol for ASD
We recommend acquiring 3D images of ASD from the upper esophagus, mid-esophagus, shallow transgastric, and deep transgastric positions. At the upper esophagus position, obtain a basal transverse view at 0° to 15° array orientation and basal short-axis view at 30° to 60°. 3D zoom mode is quite useful in this position as it provides right-sided and left-sided en face views of the atrial septum. Live 3D is used to assess device deployment in real time. At the mid-esophagus level obtain a four-chamber view at 0° to 20° array orientation and capture a full-volume acquisition for quantitation of 3D ventricular volumes. The coronary sinus is usually seen best at this position by using slight transducer retro-flexion. Also in this position a bi-caval view with 90° to 120° array orientation is used to obtain additional right-sided and left-sided en face views of the atrial septum. This view is often best for analysis of the cavo-atrial junction. At the shallow transgastric views obtain short-axis and long-axis views of the ventricles and ventricular septum in order to assess ventricular size, septal contour, and ventricular function.
Deep transgastric views are accomplished in four steps: from a standard mid-esophagus four-chamber view advance the transducer 5 to 15 cm depending on patient size, anteflex 90°, twist the transducer shaft clockwise approximately 1/4 to 1/2 turn and slowly withdraw until a deep transgastric view appears, usually the view of the left ventricular outflow tract (LVOT). Once the LVOT view is obtained spin the transducer array to 120° and twist the transducer shaft clockwise to demonstrate the atrial septum. A sagittal bicaval view obtained from the deep transgastric position with a transducer array orientation of 100° to 120° is quite useful, especially to demonstrate the inferior rim, vertical dimension of the ASD, superior rim, and the length of the atrial septum. Spinning the transducer array back to 0° to 20° and slightly deeper insertion will result in a four-chamber view. Additional or modified images may be needed in some cases in order to provide optimal delineation of the ASD and surrounding structures.
If time is limited we concentrate on the real-time live 3D modalities using 3D zoom mode to acquire en face views of the atrial septum from either the basal short-axis or bicaval view and monitor device deployment from the deep transgastric bicaval view, basal short-axis view, or mid-esophageal bicaval view. 3D full-volume views are obtained in each case but given the speed at which interventional and surgical procedures progress they are often cropped at a later time. If time allows we use all four different anatomic 3D echocardiographic modes, including live 3D, 3D zoom, full-volume 3D, and X-plane imaging. We also apply 3D color flow analysis in multiple views.
Best 3D TEE Views of ASD
Demonstration of the important anatomic features of ASD is best achieved in the following views. The RA en face view provides clear anatomic visualization of the type of ASD present, the size of the coronary sinus, and demonstrates all ASD rims simultaneously. The LA en face view demonstrates the type of ASD present, the right pulmonary vein, posterior rim, and phasic changes in ASD size and shape. A transgastric sagittal view often best demonstrates the inferior and superior rims, caval veins, septal length, and device configuration and orientation during deployment. The posterior view best demonstrates the changing alignment of the septum primum and septum secundum over the cardiac cycle, the curvilinear anatomy of the atrial septum, and the caval vein over-ride of the ASD in sinus venosus-type ASD.
Ventricular Septal Defect
Types of VSD
3D Echo of VSD Historical Perspective
The reconstruction era of 3D echo of VSD, during which 3D images were derived from reconstructed 2D images, spanned from 1994  through 2004 [15–19]. Despite the time-consuming and somewhat laborious process, and the limited image quality, investigators found that the type, location, size, spatial relationships, dimensions, and dynamic changes could all be accurately demonstrated as compared to 2D echo and intraoperative inspection. The real-time transthoracic 3D echo of VSD era began in 2004  with the initial report of images acquired with 3D full-volume and live 3D modes, followed by subsequent reports from 2006 and 2007 [21–26]. These studies included children and adults, as well as 3D full-volume mode in all and live 3D mode in some of the reports. Image quality, speed, and ease of processing were significantly improved. Additional 3D echo modalities and applications for 3D echo analysis of VSD have been described, including the use of X-plane imaging to analyze VSD in the fetus , quantitation of flow through VSD using 3D echo measurement of VSD shunt vena contracta , and the use of epicardial 3D echo of VSD during trans-ventricular VSD closure [29, 30].
Contrary to the situation with ASD and to the best of our knowledge, a detailed formal study of the best techniques and protocols to perform real-time 3D TEE of VSD has not yet been reported. Therefore, we provide a few examples of 3D TEE of VSD associated with more complex congenital cardiac anomalies in the figures of this report (see Figs. 14, 15, 16, 17 and 18).
Goals of 3D Echo of VSD
The goals of 3D imaging of VSD are to define the VSD type, position, orientation, size, dynamic changes, number of orifices, assess important surrounding structures, guide intervention, and assess residual defects.
3D Echo Imaging Recommendations for VSD
Pending completion of a formal study comparing modes, transducer positions, and other details of 3D echo methods to analyze VSD, we suggest the following approach. In our experience it has been successful in general.
Transthoracic 3D echo of VSD is usually best performed from the apical four-chamber and long axis, subcostal sagittal and coronal, and parasternal long-axis views. Full-volume 3D and live 3D acquisitions should be performed in all. 3D color Doppler flow mapping is performed in the views in which the VSD is best demonstrated. X-plane anatomic and color Doppler flow mapping is performed with the guide line of the orthogonal image placed through the plane of the VSD.
Transesophageal 3D echo of VSD is usually best performed from the mid-esophagus using the four-chamber, long-axis, and short-axis views. From the transgastric transducer positions the transgastric sagittal, coronal, and four-chamber views are most useful.
Best 3D TEE Views of VSD
Analogous to imaging an ASD, en face 3D echo views are generally quite useful for analyzing VSD morphology. The 3D anatomic depth of 3D imaging modes added to the standard 2D protocol views enhances the visualization of surrounding structures. Frontal views are also quite useful, in particular for perimembranous, muscular, and inlet VSD. The long-axis and sagittal views are often the best to demonstrate malalignment and outlet VSD.
3D echo is feasible with high diagnostic accuracy and provides additional unique detailed anatomic information. The current version of 3D echo has improved markedly over prior technologies, demonstrating significant progress over the last 5 years. Image quality, resolution, frame rate, ease of application, quantitative capabilities, and the development of improved transducers including transesophageal transducers continue to improve. 3D visualization of surfaces, volume, dynamic changes, surrounding anatomy, unique views, and advanced quantitation can now be performed rapidly and accurately. These developments overcome many of the barriers which have limited the clinical application and advancement from 2D to 3D echo imaging. However, some technical limitations persist despite extensive technical improvements, further experience must be gained, and protocols and training must be developed and implemented before 3D echo can achieve more widespread application for congenital heart anomalies. Ideal 3D echo technical features would include live 3D cropping, high resolution, high volume rate, enhanced imaging depth, live color Doppler flow analysis, single frame full-volume acquisition, and live quantitation. Hopefully these developments will all be available soon.
No potential conflicts of interest relevant to this article were reported.