If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The accuracy of two-dimensional transesophageal echocardiography (2D-TEE) for the measurement of aortic valve area (AVA) in patients with aortic stenosis (AS) depends upon the cross-section selected for imaging. Real-time three-dimensional transesophageal echocardiography (3D-TEE) may overcome this limitation of 2D-TEE. The goal of this study was to compare 3D-TEE with 2D-TEE for the measurement of AVA.
Methods and results
Twenty-five patients with AS underwent TEE. In 2D-TEE, the aortic valve image was obtained at the orifice level in the short-axis view, and AVA was measured by planimetry of the acquired images (2D-AVA). In 3D-TEE, 3D data containing the entire aortic valve were obtained. Then, a short-axis cross-section containing the smallest orifice in mid-systole was cut from the 3D data during image postprocessing, and the AVA was measured by planimetry (3D-AVA). The 3D-AVA was significantly smaller than the 2D-AVA (0.79 ± 0.35 cm2 vs. 0.93 ± 0.40 cm2, p < 0.0001), but there was a strong correlation between 3D-AVA and 2D-AVA (R = 0.94). Although the frame rate was lower in 3D-TEE than in 2D-TEE (17 ± 6 Hz vs. 58 ± 16 Hz), the 3D-AVA determined at each frame during systole showed that the difference between 3D-AVA and 2D-AVA was not explained by the lower frame rate. The time required for image acquisition of the aortic valve was shorter with 3D-TEE than with 2D-TEE (p = 0.0005).
Conclusions
The geometric AVA is smaller with 3D-TEE than with 2D-TEE, and the difference is not due to the lower frame rate of 3D-TEE. The improved accuracy of 3D-TEE along with reduced image acquisition time indicates that 3D-TEE is superior to 2D-TEE for the assessment of AVA.
The prevalence of aortic stenosis (AS) is increasing with aging of the population. Furthermore, surgery rather than medical therapy is more effective for relieving symptoms and increasing life expectancy in severe AS [
]. Accurate measurement of the aortic valve area (AVA) is important for assessing the severity of AS and the selection of an appropriate management strategy in patients with AS [
American College of Cardiology/American Heart Association Task Force on Practice Guidelines
Society of Cardiovascular Anesthesiologists
Society for Cardiovascular Angiography and Interventions
Society of Thoracic Surgeons
Bonow R.O.
Carabello B.A.
Kanu C.
de Leon Jr., A.C.
Faxon D.P.
Freed M.D.
Gaasch W.H.
Lytle B.W.
Nishimura R.A.
O’Gara P.T.
O’Rourke R.A.
et al.
ACC/AHA guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines.
]. Currently, transthoracic echocardiography (TTE) is typically used to measure AVA with the Doppler-derived continuity equation or the planimetry method. Two-dimensional transesophageal echocardiography (2D-TEE) has also been used to assess AVA by planimetry, and gives more reliable images with less noise than TTE [
]. However, 2D-TEE can overestimate the geometric AVA because of the suboptimal selection of an imaging cross-section that misses the minimal orifice area [
Quantification of the aortic valve area in three-dimensional echocardiographic data sets: analysis of orifice overestimation resulting from suboptimal cut-plane selection.
Real-time three-dimensional transesophageal echocardiography (3D-TEE) is a novel echocardiographic technique that has the potential to overcome the limitations of 2D-TEE in measuring geometric AVA, because it allows selection of a 2D cross-section from the 3D data that is independent of the ultrasound-beam direction. 3D-TEE seems to be a promising tool to assess AVA, although its accuracy and usefulness have not been fully established [
]. A recent study has shown that the geometric AVA measured with 3D-TEE is smaller than that measured with 2D-TEE, suggesting that 3D-TEE may be more accurate [
Pitfalls of anatomical aortic valve area measurements using two-dimensional transesophageal echocardiography and the potential of three-dimensional transesophageal echocardiography.
]. However, the relatively lower frame rate of 3D-TEE compared with 2D-TEE might also explain the smaller AVA, and the effect of a lower frame rate with 3D-TEE on AVA has not been clarified. Moreover, 3D-TEE may require less time than 2D-TEE to obtain the images needed for AVA assessment, but there are no reports that directly compared the time required to the measure AVA with these two methods.
The purpose of this study was to: (1) compare the geometric AVA obtained using 3D-TEE with that obtained using 2D-TEE in AS patients; (2) assess AVA per frame with 3D-TEE to evaluate the effects of a lower frame rate with 3D-TEE on AVA assessment; and (3) compare the time required for image acquisition or assessment with the two methods.
Methods
Study population
Twenty-five consecutive patients referred to our echocardiographic laboratory for decision-making of surgical treatment for AS were enrolled. All the patients were in sinus rhythm, and patients with symptoms of New York Heart Association class IV were excluded. The study protocol was approved by the Institutional Review Board of the Osaka City General Hospital and written informed consent was obtained from all patients.
Echocardiography
Conventional TTE was performed at first, and then TEE was performed immediately after. TTE and TEE were performed using a Philips IE33 instrument with S5-1 probe for TTE and X7-2t probe for TEE (Philips Medical Systems, Bothell, WA, USA). The X7-2t is a matrix array transducer that is used to acquire a pyramidal volume dataset from a single window in real time and provides 2D and live 3D images.
TTE
In TTE, standard measurements including left ventricular (LV) mass index, LV diastolic dimension (LVDd), LV systolic dimension (LVDs), LV ejection fraction (LVEF), and left atrial volume index were performed according to the Guidelines of the American Society of Echocardiography [
Recommendations for chamber quantification: a report from the American Society of Echocardiography's Guidelines and Standards Committee and Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology.
]. In addition, Doppler flow data in AS patients were obtained from the left ventricular outflow tract (LVOT) region in the pulsed wave mode and from the aortic valve in the continuous wave mode using multiple transducer positions to obtain the maximum velocity. The LVOT diameter was measured in the parasternal long-axis view at the position used to obtain the pulse-wave Doppler data. AVA was calculated using the continuity equation [AVA = areaLVOT × (velocity time integralLVOT/velocity time integralvalve)] [
Insertion of the TEE probe was performed using standard methods. After probe insertion, the transmitting and receiving frequencies were set at 2–7 MHz for fundamental imaging and 2.4/4.8 MHz or 2.7/5.4 MHz for 2nd harmonic imaging, with the goal of obtaining higher resolution aortic valve images with fewer artifacts. The gain, compression and focus depth were also adjusted as appropriate.
Next, a 2D-TEE motion picture of the aortic valve was recorded. The long-axis view of the aortic valve and ascending aorta was obtained initially to define the optimal position of the transducer location at the mid-esophageal level. Holding the transducer position stable, the imaging plane was rotated to obtain the short-axis view. Then, the imaging sector was narrowed to image the aortic valve alone and obtain the maximal frame rate. A 2D cross-section in the short-axis view was carefully selected so as not to miss the minimal orifice area at the tip of the leaflets at the time of maximal opening in mid-systole, and a motion picture of the aortic valve during one complete cardiac cycle was digitally recorded. The frame rate was also recorded.
Finally, a 3D-TEE motion picture of the aortic valve was recorded. Using the ‘live 3D’ function, perpendicular two cross-sectional 2D-images of the aortic valve in long-axis and short-axis views were displayed side by side. Adjusting the imaging sector containing the entire aortic valve in each image during one complete cardiac cycle to obtain the maximal frame rate and not to miss any parts of the aortic valve, a real-time 3D volume dataset was acquired and the frame rate was recorded (Fig. 1).
Figure 1Real-time 3D data acquisition including the entire aortic valve using the “live 3D” function.
After the completion of the TEE examination, 2D-AVA and 3D-AVA were measured as part of the postprocessing analysis. For the measurement of 2D-AVA, the frame with the maximal opening of the aortic valve in mid-systole was identified. Then, the contours of the inner cusps were traced, and the area was measured by planimetry. For the measurement of 3D-AVA, three cross-sectional 2D-images perpendicular to each other (x, y, and z planes) were obtained from the 3D-volume dataset using the built-in commercial software (QLAB, Philips Medical Systems). With this software, the 3D volume image was obtained at mid-systole, where there was maximal opening of the aortic valve. Then, 3 perpendicular 2D planes were cut from the 3D volume image, and the two long-axis planes that were perpendicular to each other were adjusted to cross the center of the aortic valve. The third cut-plane showing the short-axis view of the aortic valve was tilted and parallel shifted at the tip of the leaflets to select the minimal orifice area. Using the image planes that gave the minimal orifice area, the contours of the inner cusps were traced, and the 3D-AVA was measured by planimetry (Fig. 2).
Figure 2Three cross-sectional 2D-images obtained from the 3D-data with QLAB software. (A) Two long axes perpendicular to each other were set across the center of the aortic valve (blue line in the left upper panel and green line in the left lower panel). The 2D cross-sectional image at the blue line is displayed in the left lower panel, and 2D cross-sectional image at the green line is displayed in the left upper panel. Then, an optimal short-axis cut-plane was selected by shifting and tilting the red lines at the left upper and lower panels to acquire the minimal orifice area at the time of maximal opening in mid-systole. The short-axis cross-sectional image at the red lines is displayed in the right upper panel, and 3D-aortic valve area (AVA) was measured with planimetry. (B) Suboptimal cut-plane parallel shifting to the base of the valve leaflets (see a bold red line). AVA should be overestimated at this cross-section. (C) Suboptimal cross-section parallel shifted towards the leaflet tips. The edge of valve leaflets to measure AVA is partly broken at this cross-section, and it can be recognized that the cross-section has been shifted towards the leaflet tips over the optimal level (see bold red lines).
In addition to obtaining the minimal 3D-AVA, we used planimetry to measure the AVA from all the 3D-TEE frames recorded during one systole phase in each patient. The selection of the cross-section with the minimal orifice area was repeated in each frame in this analysis. Then, the 3D-AVA values were plotted against the frame time during systole, and then the frames from different patients were aligned so that the maximum AVA occurred at the same relative time in each patient. The 3D-AVA values obtained for each frame were compared with the 2D-AVA.
The times required for aortic valve image acquisition during 2D-TEE and 3D-TEE were also measured. In addition, we measured the times required for 2D-TEE and 3D-TEE image postprocessing for measurement of the AVA by planimetry.
These postprocessing analyses of 2D-AVA and 3D-AVA were performed by one expert cardiologist twice to assess intraobserver variability and by another expert cardiologist to assess interobserver variability.
Statistical analysis
Continuous variables are expressed as mean ± standard deviation (SD). Comparison of measurements between two groups was performed using Student's paired t test. Linear regression analysis was used to assess the relationship between 2D-AVA and 3D-AVA, and correlation coefficients were expressed as R values. Bland–Altman analysis was performed to assess the agreement between two methods, and the limits of agreement were defined as mean ± 1.96 SD of the average differences between the methods. Intraobserver and interobserver variabilities were assessed as correlation coefficients (R) with linear regression analysis and mean differences. A p-value < 0.05 was considered statistically significant.
Results
The study population consisted of 25 patients with AS. There were no particular complications during any of the TTE or TEE examinations. The image acquisitions with 2D-TEE and 3D-TEE were feasible in all the patients. Table 1 shows clinical characteristics of the 25 patients. There were 12 men and 13 women, with a mean age of 71 ± 8 years. Table 2 shows the standard measurements obtained with TTE. The mean LV mass index was 138 ± 35 g/m2. The mean LVEF was 58 ± 13%, and 6 patients (24%) had LV systolic dysfunction (LVEF < 50%). Twenty (80%) had a degenerative calcified valve, 3 (12%) had a bicuspid valve, and 2 (8%) had a rheumatic valve. The average peak and mean pressure gradients across the aortic valve were 75 ± 28 mmHg and 45 ± 19 mmHg, respectively. The mean AVA derived from the continuity equation was 0.75 ± 0.20 cm2.
The averages of 3D-AVA and 2D-AVA were 0.79 ± 0.35 cm2 and 0.93 ± 0.40 cm2. In addition, 3D-AVA showed a very strong correlation with 2D-AVA (R = 0.94, p < 0.0001) (Fig. 3A) . The limits of agreement between the two methods used to measure AVA were evaluated with a Bland–Altman analysis. The mean of the individual difference between 3D-AVA and 2D-AVA values was −0.14 ± 0.27 cm2 (Fig. 3B).
Figure 3(A) Linear regression analysis between 3D-aortic valve area (AVA) and 2D-AVA. (B) Bland–Altman analysis between 3D-AVA and 2D-AVA. The graph displays the difference of the individual values (y-axis) against the average values (x-axis) between 3D-AVA and 2D-AVA. The central line represents the mean of the differences between the individual values and the upper and lower lines represent the limits of agreement.
Fig. 4 shows the 3D-AVA values plotted against the frame time for one systolic phase. The 2D-AVA is also shown in Fig. 4 for comparison. The frame rate of 3D-TEE was significantly lower than that of 2D-TEE (17 ± 6 vs. 58 ± 16 Hz, p < 0.0001). The time for image acquisition in 3D-TEE was significantly shorter than that in 2D-TEE (45 ± 34 vs. 85 ± 41 s, p = 0.0005), whereas the time for image postprocessing was longer with 3D-AVA than with 2D-AVA (202 ± 60 vs. 66 ± 71 s, p < 0.0001) (Table 3).
Figure 4The 3D-aortic valve area (AVA) values are plotted against frame time during systole, and then the frames from different patients were aligned so that the maximum AVA occurred at the same relative time in each patient.
The intraobserver variabilities as assessed by the correlation coefficient and the mean difference were 0.94 and 0.09 ± 0.08 for 3D-AVA and 0.95 and 0.11 ± 0.07 cm2 for 2D-AVA. The interobserver variabilities were 0.96 and 0.08 ± 0.08 cm2 for 3D-AVA, and 0.92 and 0.14 ± 0.11 cm2 for 2D-AVA.
Discussion
Assessing the severity of AS is a commonly encountered clinical problem and is limited by the accuracy of the current methods including TTE and TEE. Our goal in the present study was to assess the feasibility and usefulness of newly available real-time 3D-TEE for the measurement of AVA. TEE with image acquisition for 3D reconstruction was successfully performed in all patients in the present study.
In the present study, 3D-AVA and 2D-AVA had a good correlation with each other. However, 3D-AVA was significantly smaller than 2D-AVA. We thought that this difference between 3D-AVA and 2D-AVA occurred due to the following limitations in the measurement of 2D-AVA. 2D-TEE often makes an oblique cut or cut-plane through the valve leaflets rather than at the tip of the leaflets. Consequently, it is often technically difficult to capture the tip of the aortic valve leaflets at their maximal systolic opening using 2D-TEE with a fixed imaging plane [
Quantification of the aortic valve area in three-dimensional echocardiographic data sets: analysis of orifice overestimation resulting from suboptimal cut-plane selection.
Three-dimensional surface area of the aortic valve orifice by three-dimensional echocardiography: clinical validation of a novel index for assessment of aortic stenosis.
Comparison of accuracy of aortic valve area assessment in aortic stenosis by real time three-dimensional echocardiography in biplane mode versus two-dimensional transthoracic and transesophageal echocardiography.
]. In addition, 2D-TEE has the disadvantages of more fussy contours, shadow phenomena, and motion artifacts; and the gain setting may affect the measured AVA, especially in severely calcified valves [
Quantification of the aortic valve area in three-dimensional echocardiographic data sets: analysis of orifice overestimation resulting from suboptimal cut-plane selection.
]. Because of these limitations, it is difficult to use 2D-TEE to measure the minimal orifice area of the aortic valve when it is maximally opened, and it is easy to overestimate AVA with 2D-TEE. In contrast, 3D-TEE can cut a 2D cross-section from the 3D data that is independent of the ultrasound-beam direction, and the cut-plane can be shifted and angled to obtain an image at the tip of the leaflets. Thus, 3D-TEE may overcome one of the limitations of the measurement of geometric AVA. Several previous studies suggested that 3D-TEE provides a more accurate orifice area and overcomes the limitations of 2D-TEE for measuring AVA [
Quantification of the aortic valve area in three-dimensional echocardiographic data sets: analysis of orifice overestimation resulting from suboptimal cut-plane selection.
Three-dimensional surface area of the aortic valve orifice by three-dimensional echocardiography: clinical validation of a novel index for assessment of aortic stenosis.
Comparison of accuracy of aortic valve area assessment in aortic stenosis by real time three-dimensional echocardiography in biplane mode versus two-dimensional transthoracic and transesophageal echocardiography.
Determination of aortic valve area in valvular aortic stenosis by direct measurement using intracardiac echocardiography: a comparison with the Gorlin and continuity equations.
]. However, these previous studies used biplane or multiplane 3D-TEE instruments, which did not have a real-time 3D mode and needed several recordings at each angle during different cardiac cycles to construct a single 3D image. In contrast, we used real-time 3D-TEE in the present study, and could record a 3D volume dataset during a single cardiac cycle. One recent study has also assessed AVA using real-time 3D-TEE, and found results consistent with the present study [
]. They measured AVA by tracing the aortic valve orifice with thin-thickness 3D images. Their method seems to have an advantage for more distinct delineation of the edge of the aortic valve orifice. In contrast, we used a 2D plane cut from the 3D volume data, and our method seems to have an advantage in selecting the cross-section with the minimal orifice area. It is uncertain which method is more accurate in the assessment of AVA, and additional studies are required to address this issue.
Because of the lower frame rate of real-time 3D-TEE, it is possible that the frame with maximal opening of the aortic valve was missed, and this might have resulted in a reduced AVA compared with 2D-TEE [
Pitfalls of anatomical aortic valve area measurements using two-dimensional transesophageal echocardiography and the potential of three-dimensional transesophageal echocardiography.
]. To address this issue, we measured the AVA in each 3D-TEE frame in all patients and compared each of these measurements with 2D-AVA. The 3D-AVA values plotted against frame time during one systolic phase suggested that the lower frame rate was not the cause of the reduced 3D-AVA compared with the 2D-AVA.
The analyses of reproducibility in the measurements of AVA in the present study showed strong intraobserver and interobserver agreements both in 2D-AVA and 3D-AVA. This suggested that the differences between 2D-AVA and 3D-AVA seen in the present study did not depend on the differences in the intraobserver or interobserver variabilities.
Our results showed that the time required for image acquisition was longer with 2D-TEE than with 3D-TEE. This is also due to the difficulty in obtaining the optimal cut plane of the orifice area in 2D-TEE. In contrast, only a single dataset that included the entire aortic valve was needed for the 3D-TEE analysis. Therefore, 3D-TEE would shorten the examination time and possibly minimize patients’ distress. In contrast, the postprocessing time required for AVA measurement was shorter with 2D-TEE than with 3D-TEE. This resulted from the additional time required with 3D-TEE to cut out an optimal 2D-plane from the 3D volume dataset. However, we believe that this is an advantage rather than a disadvantage of 3D-TEE. In 3D-TEE, we can repeatedly cut a 2D-plane from the 3D data until an optimal 2D image is obtained. Therefore, 3D-TEE can generate more careful and accurate measurements during image postprocessing.
Limitations
Some limitations of the present study should be acknowledged. First, the number of subjects in the present study was small; and the etiology of AS was not the same among patients, since we included patients with degenerative calcified, bicuspid, and rheumatic valves.
Second, we did not perform catheter-based measurements of AVA using Gorlin's formula, which is a gold standard for the assessment of AVA. However, catheter-derived AVA is the physiological AVA and differs from the geometric AVA that we investigated in the present study. In addition, it has a drawback known as the pressure recovery theory [
]. Physiological AVA can also be calculated with continuity equation using Doppler echocardiography. However, it is theoretically smaller than geometric AVA [
Differences in aortic valve area with CT planimetry and echocardiography (continuity equation) are related to divergent estimates of left ventricular tract area.
Differences in aortic valve area with CT planimetry and echocardiography (continuity equation) are related to divergent estimates of left ventricular tract area.
]. For these reasons, we believe that 3D-AVA has the potential to become a gold standard to assess AVA.
Conclusion
Geometric AVA measured with 3D-TEE is smaller than that measured with 2D-TEE, and the difference is not due to the lower frame rate of 3D-TEE. The improved accuracy of 3D-TEE along with reduced image acquisition time indicates that 3D-TEE is superior to 2D-TEE for the assessment of AVA.
References
Stewart B.F.
Siscovick D.
Lind B.K.
Gardin J.M.
Gottodiener J.S.
Smith V.E.
Kitzman D.W.
Otto C.M.
Clinical factors associated with calcific aortic valve disease.
American College of Cardiology/American Heart Association Task Force on Practice Guidelines
Society of Cardiovascular Anesthesiologists
Society for Cardiovascular Angiography and Interventions
Society of Thoracic Surgeons
Bonow R.O.
Carabello B.A.
Kanu C.
de Leon Jr., A.C.
Faxon D.P.
Freed M.D.
Gaasch W.H.
Lytle B.W.
Nishimura R.A.
O’Gara P.T.
O’Rourke R.A.
et al.
ACC/AHA guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines.
Quantification of the aortic valve area in three-dimensional echocardiographic data sets: analysis of orifice overestimation resulting from suboptimal cut-plane selection.
Pitfalls of anatomical aortic valve area measurements using two-dimensional transesophageal echocardiography and the potential of three-dimensional transesophageal echocardiography.
Recommendations for chamber quantification: a report from the American Society of Echocardiography's Guidelines and Standards Committee and Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology.
Three-dimensional surface area of the aortic valve orifice by three-dimensional echocardiography: clinical validation of a novel index for assessment of aortic stenosis.
Comparison of accuracy of aortic valve area assessment in aortic stenosis by real time three-dimensional echocardiography in biplane mode versus two-dimensional transthoracic and transesophageal echocardiography.
Determination of aortic valve area in valvular aortic stenosis by direct measurement using intracardiac echocardiography: a comparison with the Gorlin and continuity equations.
Differences in aortic valve area with CT planimetry and echocardiography (continuity equation) are related to divergent estimates of left ventricular tract area.
00017-2&id=gr1.jpg){kind=link}
00017-2&id=gr2.jpg){kind=link}
00017-2&id=gr3.jpg){kind=link}
00017-2&id=gr4.jpg){kind=link}