There has been a considerable evolution in technology since the initial description of 3D echocardiography several decades ago.1 The first generation of 3D probes performed gated acquisition with static rendering of images, but not in realtime. The second generation did to an extent allow realtime imaging, but the relatively low density of active elements (approximately 300) resulted in a low sampling rate and compromise in terms of image quality. Current 3D probes, due to the high element count (>3,000), deliver the high sampling rates necessary to achieve clinically useful realtime 3D imaging with high image resolution and satisfactory frame rates.
Imaging in 3D essentially involves the acquisition of a volume data set over one or more heartbeats, depending on volume size. In ‘full-volume’ acquisitions, typically four or more sub-volumes are acquired with electrocardiogram (ECG) gating and are stitched together to create a larger volume. This is used for larger structures, for example in the assessment of left ventricular function and volumes. Full-volume acquisition can also be performed with colour Doppler to assess the origin and shapes of flow jets in 3D. This often requires more sub-volumes and ECG cycles (typically seven), and the total volume acquired is often smaller than the full volume without colour Doppler. As might be expected, the major limitation of this approach is the creation of ‘stitching’ artefacts when the sub-volumes are not aligned, e.g. irregular heart rhythm such as ectopic beats or atrial fibrillation and with significant translational motion of the heart such as respiration. Single-heartbeat large-volume acquisitions are under development; however, a limitation of the current technology is that as the volume of interest increases in size, image quality is compromised in terms of spatial and temporal resolution. If a smaller 3D volume is obtained, ‘live’ (or realtime) 3D imaging with good resolution and frame rates is possible. This is useful for studying a single structure of interest such as the mitral valve, and removes the problem of irregular heart rhythms. However, colour Doppler is currently not available with live 3D imaging. As with all echocardiography techniques, 3D methods are still dependent on obtaining an adequate echo window.
Measurements can be made of distance or area from any 2D plane within the 3D data set, but direct measurements from the 3D image display are currently not possible, although an appreciation of size can be made by overlaying a grid of known dimensions.
The time taken for analysis previously limited its clinical utility: not only were acquisition times long, at five minutes, but 3D image reconstruction and analysis often took a further 20 minutes. This was further hampered by artefacts produced in the data set from irregular cardiac cycles and respiration. The first realtime probes to become available were capable of transthoracic 3D (3D transthoracic echocardiography [TTE]) imaging but had limitations due to suboptimal image display. With the introduction of transeosphageal 3D (3D transeosphageal echocardiography [TEE]), realtime imaging, probe image quality, acquisition and analysis of the data sets can be performed online within seconds.
Clinical Applications of 3D Echocardiography
Imaging in 3D provides clear appreciation of the true shape of cardiac structures and the spatial relationships between them. Full evaluation of valvular heart disease, whether for diagnosis or for planning of an intervention, requires accurate delineation of the actual valve pathology but also assessment of any consequent cardiac dysfunction, including assessment of ventricular size and function. 3D methods can also be used for assessment of ventricular dyssynchrony and for morphological assessment in congenital lesions, but these applications will not be discussed further in this review. Table 1 shows common clinical applications of 3D TTE and 3D TEE. The clinical utility of 3D echocardiography in valvular disease is well illustrated by its use in mitral valve disease, which will be covered in more detail below.
What Is the Role of 3D Echocardiography in the Assessment of Mitral Regurgitation?
The role of 3D echocardiography in the assessment of mitral regurgitation is best understood in terms of the assessment of valve morphology (aetiology and mechanism) and then assessment of the severity of regurgitation.
Understanding Mitral Valve Morphology
Mitral regurgitation can arise from a variety of mechanisms that may occur in combination. Traditionally described by Carpentier’s classification based on leaflet motion (normal motion, prolapse or restriction of a leaflet segment), the location and extent of leaflet motion abnormality guides treatment options. Guidelines emphasise the benefits of valve preservation by repair on patient outcomes.2 An accurate description of the mechanism of valve failure is required to predict the complexity of the techniques needed and hence the likelihood of achieving a successful mitral repair.
This helps to guide the decision on the best time to intervene and offer surgery; this has become increasingly relevant as mitral repair is being performed in asymptomatic patients with severe mitral regurgitation.3,4 However, if there is a high chance that the valve may be replaced, surgery is better delayed until symptoms develop or other parameters are met, such as a dilating ventricle or falling left ventricular ejection fraction.
Transthoracic 3D Echocardiography in the Assessment of Mitral Valve Morphology
Assessment of the mitral valve by 2D echocardiography requires the operator to obtain multiple views through all segments of the two leaflets. This requires considerable expertise and experience, but even then errors in interpretation may occur.
Surgical View of the Mitral Valve
3D TTE can provide a rapid overview of the mitral valve using the en face or ‘surgical’ view from the left atrium. This is unobtainable with the 2D approach. Realtime 3D (live 3D or ‘zoom’ mode) is performed preferably from the parasternal window in either the long or short axis. The resolution is optimal from these windows since the mitral valve is in close proximity to the probe. The live image is rotated to view the left atrial aspect of the mitral valve and the image is optimised by gentle probe angulation to bring the leaflets, coaptation line and annulus into view in their entirety. Visualising the heart from the transthoracic approach means one of the main limitations is image quality. Frame rates and image resolution are the main factors that limit precise details of valve anatomy. However, in the majority of patients at least adequate image quality allows a useful and rapid overview of the valve morphology. The extent of any prolapsing or restricted leaflet segments, the commissures, leaflet coaptation line and mitral annulus can be seen.
Segmental Analysis of the Mitral Valve
3D echocardiography also provides the ability to perform accurate segment-by-segment analysis.5 The preferred format is full-volume acquisition, as frame rates and image resolution are greatest. Again, the parasternal long-axis window is preferred as anatomical landmarks using standardised analysis protocols mean this approach is simple and reproducible;6 where such windows are poor, the apical windows can be used.
Rapid online segmental analysis can be performed by viewing the data set in a multiplanar reconstruction format to define short- and long-axis views of the mitral valve. Using established landmarks, the entire coaptation line can be assessed from the anterolateral commissure, P1 and A1 segments through P2, A2 and then P3, A3 and the posteromedial commissure. The relationship of each segment to the annulus can be accurately assessed. Measurements can be made of segment prolapse or tenting7 in relation to the annulus. Acquisition times and the time needed for analysis of the data sets are reasonable and can easily be incorporated into the standard 2D echocardiography examination. Figure 1 gives an example of segmental analysis.
The literature confirms the clinical utility of 3D echocardiography in mitral valve morphology diagnosis. Comparison of real-time 3D TTE with 2D TEE has shown comparable high sensitivity, specificity and accuracy for identifying prolapsing mitral segments;8 a small proportion (11%) of patients were excluded due to insufficient image quality by 3D TEE. Pepi et al.,9 in a series of 102 patients, showed that the time needed to obtain and analyse the 3D TTE images was 7±4 minutes. The quality of 3D TTE was insufficient in 8% of cases, sufficient in 16%, good in 55% and optimal in 21%. They performed a comprehensive 3D protocol taking views from both parasternal and apical windows. A more focused approach with the acquisition of two 3D images (a realtime 3D image and a full-volume data set) can reduce these times further. Pepi et al.9 found similar diagnostic accuracies for 2D TEE versus 3D TTE (87 and 90%, respectively; p=NS). Others have found similar comparable diagnostic data between 2D TEE and 3D TTE.5,10,11
Transoesophageal 3D Echocardiography in the Assessment of Mitral Valve Morphology
With the huge step up in image resolution offered by 3D TEE, an authentic depiction of mitral valve anatomy can now be viewed in vivo and in realtime. New levels of understanding of valve morphology are now possible.
Surgical View of the Mitral Valve
Since the closest cardiac structure to the probe in TEE is the left atrium, replicating the surgical view is rapid and straightforward. Using clear 2D image quality as the guide and choosing the image plane where the mitral valve lies as perpendicularly to the plane as possible, the surgical view is easily obtained in live 3D zoom mode.
Figure 2 shows examples of posterior mitral leaflet prolapse on 3D TEE and illustrates how this allows appreciation of differences between them. The huge advantage of this technique is its ability to produce excellent image quality irrespective of heart rhythm. Additional information can be gained by 3D colour full volume imaging (but image quality may be reduced in the setting of arrythmias), and with complex lesions can aid in better understanding of the mechanism of valve failure.12 Not only can the location of a primary regurgitant jet be identified, but additional smaller jets along the coaptation line and commissures may be seen. These can play an important role in residual regurgitation post-repair and may help draw the surgeon’s attention to areas of the mitral leaflet coaptation requiring attention.
Segmental Analysis of the Mitral Valve
The same approach described for 3D TTE segmental analysis can be applied to 3D TEE. Since the image resolution is superior, this form of analysis can be performed on the same live 3D zoom mode data set acquired for the surgical view. Very little extra time is needed to obtain this 3D view and this can easily be incorporated into the routine 2D TEE study (aquisition time 60±18 seconds).13
Macnab et al. compared 2D and 3D TEE assessment of the mitral valve against the gold standard of surgical findings, and reported that 3D TEE is more accurate at identifying the location and extent of leaflet prolapse, and was more so at the most medial and lateral portions of the valve.14 More recently, Grewal et al. performed a similar study and also found superior performance of 3D TEE, particularly in identifying complex disease involving multiple segments or both leaflets.13 3D TEE is superior to 2D TEE and 3D TTE in its diagnostic accuracy of localisation of leaflet prolapse (96.5, 87 and 90%, respectively; p<0.001). The advantage that 3D TEE offers over 3D TTE is the significant improvement in image resolution and the superb detail seen in each leaflet. Anatomy that was not appreciated previously, such as unusually deep clefts within the leaflet (see Figure 3) as the site of regurgitation, are being understood.
Mitral Valve Quantification Software
Mitral valve quantification has the potential to provide new insights into mitral pathophysiology concepts and may help to refine repair strategies. A recent study by Grewal et al.15 using realtime 3D TEE found that the normal mitral annulus reduction in size (mitral orifice area) in early systole and saddle-shape deepening contributes to mitral competency. In mitral valve disease these early systolic changes were less pronounced despite a similar magnitude of ventricular contraction, suggestive of ventricular–annular decoupling. Subsequent area enlargement may contribute to mitral incompetence. In patients undergoing mitral repair, the annulus remained dynamic without systolic saddle-shape accentuation. Currently, such detailed analysis is time-consuming and requires offline analysis with customised software.
However, software is available online (including MVQ, Phillips Medical and Tomtec mitral quantification tool) that allows analysis of 3D TEE data sets in clinical practice. MVQ offers a semi-automated analysis package of mitral valve anatomy at end-systole. This can provide a very detailed assessment with accurate measurements of the annulus and leaflets, papillary muscle position and aortic–mitral angles, and is likely to play an important role in pre-operative planning of mitral valve repair in the future. Currently, little published data exist on its use in clinical practice. In our experience this software provides excellent information on mitral annulus dimensions and geometry. Useful parameters include antero-posterior diameter, commissural diameter, total circumference, annulus saddle height, total mitral orifice area and mitral annulus to aortic valve angle. In complex anatomy, depiction of the leaflet segments in relation to the mitral valve annulus often clarifies an otherwise difficult assessment even with the use of segmental analysis, since such complicated anatomy can be a challenge to decipher with the naked eye (see Figure 4). In our laboratory analysis, the time taken averages approximately five to seven minutes. The ability to perform a segmental analysis post-repair and assess the degree of leaflet apposition along the whole length of coaptation, coupled with quantification of the reduction in annulus dimensions with MVQ software, is likely to provide further insights into durability of repair.
Assessment of Mitral Regurgitation Severity
Assessment of the severity of mitral regurgitation by 2D echocardiography can be challenging, particularly with eccentric jets. 3D colour Doppler allows better visualisation of the origin, size and shape of jets.16,17 However, previous 3D methods of regurgitation quantification relied on gated acquisition, and although they were shown to be more accurate than the established 2D echocardiographic methods, they were time-consuming and not clinically applicable.18,19 The vena contracta and proximal isovelocity surface area (PISA) assessment are methods used to quantify severity of mitral regurgitation, the latter allowing calculation of effective regurgitant orifice area and regurgitant volume.20 More recently, with the advent of the 3D fully sampled matrix array probes, 3D colour flow can now be displayed with the greyscale image. The conventional PISA method assumes a hemispheric shape of the isovelocity surface, which very often is not the case (see Figure 5). These newer 3D methods allow assessment of the true shape of proximal flow convergence region and permit more accurate measurement of the PISA and vena contracta.21,22 Such data have improved our understanding of the shape of the regurgitant orifice.23 In degenerative mitral valve disease the PISA is more spherical, but in functional mitral regurgitation the PISA is elongated and elliptical. These differences in PISA geometry in part explain discrepancies in regurgitation severity by 2D PISA methods. Applying a hemi-ellipsoid formula reduces the underestimation of regurgitation from 49 to 26%.24 Despite these advances, these methods are not in routine clinical use. Limitations include time-consuming analysis methods, image resolution and dependence on multiple cardiac cycles (seven to 14 cycles) with narrow sector acquisitions, which increase artefacts.
Percutaneous Catheter-based Interventions for Mitral Regurgitation and the Role of 3D Echocardiography
While surgical repair remains the therapy of choice in severely regurgitant mitral valves, functional mitral regurgitation often involves significant left ventricular impairment. Such patients may face significant risk from conventional surgery, and percutaneous approaches are currently being explored. MitraClip (Evalve Inc.) is one example. It is based on the surgical Alfieri operation and inserted using a transvenous approach and atrial septal puncture.
The ‘clip’ is positioned over the regurgitant orifice grasping both mitral leaflet free edges, creating a double orifice with significant reduction in mitral regurgitation.25 3D TEE provides excellent assessment of the mitral annulus dimensions, the location and extent of the regurgitant orifice and leaflet morphology, all of which are essential in planning the procedure. During the procedure, 3D TEE is also of considerable value.26 The position of transseptal puncture is crucial: if it is too high or too low on the atrial septum, the delivery system may either not reach the valve or assume an awkward angle, making clip deployment impossible. 3D imaging allows a precise atrial septal puncture and accurate positioning over the mitral valve leaflets during clip deployment. Once successfully deployed, assessment of residual regurgitant jets and how this relates to valve anatomy can also be performed using 3D TEE.
What Is the Role of 3D Echocardiography in the Assessment of Mitral Stenosis
Mitral stenosis severity is quantified in terms of mean transmitral gradient and mitral orifice area.27 Methods that calculate the ‘effective’ mitral orifice area include pressure half-time measurement (PHT), continuity equation and PISA methods. However, these have their limitations since the Doppler-derived methods are influenced by factors including tachycardia, heart rhythm, non-linear Doppler velocity curves and concomitant valvular disease, and are not applicable in the immediate period post-balloon valvuloplasty.
Direct planimetery can be used to measure the true ‘anatomical’ orifice area. This is considered to be the reference method in clinical practice, having been shown to correlate more precisely with direct measurement of anatomical orifice at surgery than Doppler methods.28 However, the stenotic mitral valve leaflets often form a funnel shape and the stenotic orifice may be situated obliquely within the ventricle. 2D planimetry is limited by difficulties in obtaining the minimum cross-sectional area during planimetry measurements. This method demands considerable experience and expertise to define the correct orientation of the true mitral valve orifice. 3D TTE helps locate the plane with the smallest mitral valve orifice as the image can be viewed using multiplanar reformatting to align with the orifice in both the long and the short axis (see Figure 6). Use of 3D has been shown to be an accurate method for assessing mitral valve area, and is faster and more reproducible, even in less experienced hands, than 2D echocardiography.29,30 If the patient is in atrial fibrillation, instead of the full volume, live 3D imaging can be used to overcome stitching artefacts. Invasive catheter measurements using Gorlin’s method have also been compared with 3D echocardiography, and the latter appears to correlate more closely than other non-invasive methods.31
Another limitation of 2D planimetry is the presence of significant calcification, particularly within the leaflets. This can be overcome by 3D TEE imaging, as clear imaging of the orifice is possible from the left atrial en face view as calcific shadowing is cast into the left ventricle. Alternatively, the use of 3D TTE with colour Doppler is another solution.32 It seems likely that 3D techniques will become the new ‘gold standard’ for quantifying mitral valve area.
Percutaneous balloon valvuloplasty is an effective treatment for appropriately selected patients with rheumatic mitral stenosis. 3D TEE can provide accurate information on the functional significance and morphology of the valve to determine suitability for balloon valvuloplasty. Accurate peri-procedural evaluation of the mechanism and effect of valvuloplasty is also desirable. Assessment of complications can be better appreciated with 3D imaging, as with 3D assessment one can determine the increase in valve area and whether this has occurred due to commissural splitting as intended, or due to leaflet tearing (see Figure 7). It is also recognised that Doppler-derived measures of mitral valve area are inaccurate immediately post-valvuloplasty due to acute changes in left atrial compliance and transmitral gradient.33 The most reliable method of assessing change in area is direct planimetry, and 3D TTE has been shown to correlate best with invasive measurements immediately post-procedure.34
The Place of 3D Echocardiography in Mitral Valve Disease in Clinical Practice
In clinical practice, 3D used as part of the initial TTE study could largely replace the need for initial 2D TEE as it allows more confident assessment of leaflet morphology, and hence the likelihood of successful repair and timing of intervention. Once the decision for surgery is made, a TEE with 3D imaging could be performed to accurately define valve anatomy and dimensions used for preoperative planning. In mitral stenosis, 3D planimetry is likely to become the method of choice in assessing severity of stenosis and is particulary useful immediately following balloon mitral valvuloplasty.
As with all new techniques there is a significant learning curve, and image quality is still dependent on echocardiographic windows. Once users become familiar with the technology, adding 3D acquisitions onto a standard 2D study typically requires only a few extra minutes, although more time may be required afterwards for ‘post-processing’. 3D echocardiography provides incremental information over standard 2D techniques, allowing more accurate assessment of mitral valve disease. This can be used to refine the diagnosis and better guide the treatment of patients. Current clinical applications of the technology in the assessment of mitral valve disease have been described above, although the clinical uses of 3D echocardiography will likely continue to grow in this area.