(Circulation. 1997;96:3687-3695.)
© 1997 American Heart Association, Inc.
Articles |
Direct Measurement of Three-dimensionally Reconstructed Flow Convergence Surface Area and Regurgitant Flow in Aortic Regurgitation
In Vitro and Chronic Animal Model Studies
From The Clinical Care Center for Congenital Heart Disease, Oregon Health Sciences University, Portland (T.S., X.L.,M.I., S.H., D.J.S.); the Laboratory of Animal Medicine and Surgery, NHLBI, Bethesda, Md (M.J., I.Y.); and Non-invasive Cardiac Laboratory, Tufts-New England Medical Center, Boston, Mass (A.D., P.A., N.G.P.).
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background Evaluation of flow convergence (FC) with two-dimensional (2D) imaging systems may not be sufficiently accurate to characterize these often asymmetric, complex phenomena. The aim of this study was to validate a three-dimensional (3D) method for determining the severity of aortic regurgitation (AR) in an experimental animal model.
Methods and Results In six sheep with surgically induced chronic AR, 20 hemodynamically different states were studied. Instantaneous regurgitant flow rates were obtained by aortic and pulmonary electromagnetic flow meters. Video composite data of color Doppler flow mapping images were transferred into a TomTec computer after computer-controlled 180° rotational acquisition. Direct measurement of the 3D reconstructed FC surface areas as well as measurements of FC areas estimated with 2D methods with hemispherical and hemielliptical assumptions were performed, and values were multiplied by the aliasing velocity to obtain peak regurgitant flow rates. There was better agreement between 3D and electromagnetically derived flow rates than there was between the 2D and the reference values (r=.94, y=1.0x-0.16, difference=0.02 L/min for the 3D method; r=.80, y=1.6x-0.3, difference=1.2 L/min for the 2D hemispherical method; r=.75, y=0.90x+0.2, difference=-0.20 L/min for the 2D hemielliptical method).
Conclusions Without any geometrical assumption, the 3D method provided better delineation of the FC zones and direct measurements of FC surface areas, permitting more accurate quantification of the severity of AR than the 2D methods.
Key Words: regurgitation • echocardiography • blood flow • Doppler analysis
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
A number of noninvasive methods have been reported to be useful for evaluating the severity of aortic regurgitation.1 2 3 4 5 6 7 8 9 10 11 12 However, most echo-Doppler methods such as color Doppler jet area planimetry, regurgitant flow velocity continuous-wave (CW) pressure half-time, and retrograde flow signal in the descending aorta are influenced by many instrument as well as physiological factors.2 3 5 9 Additionally, these methods are not truly quantitative.
Laminar acceleration phenomena for flows toward aortic and mitral orifices imaged by magnetic resonance imaging and color Doppler flow mapping have been studied experimentally and clinically regarding their use for quantifying regurgitant and stenotic flow rates.13 14 15 16 17 18 19 20 21 22 23 24 25 With the flow convergence methods, the flow rate through regurgitant or stenotic orifices is calculated as the product of the isovelocity surface area and its corresponding velocity based on the continuity concept.13 14 15 16 17 18 21 22 23 24 25
To accomplish this, it is essential to obtain accurate isovelocity surface areas for estimating the flow rates. Considering the three-dimensional nature of flow convergence phenomena, hemispherical or even hemielliptical assumptions of the geometry of the isovelocity surface used for previously reported flow convergence studies using two-dimensional imaging systems may be overly simplified. The shapes of the isovelocity surfaces can be quite variable, depending on the aliasing velocities selected and the surrounding geometry, as has been pointed out in the study by Schwammenthal et al25 and in our own earlier studies.17 Two-dimensional imaging methods require mental reconstruction and assumptions about shapes for the flow convergence isovelocity surface that can be problematic for flows associated with complex orifice geometry. In contrast, direct measurement of three-dimensionally reconstructed flow convergence surface areas does not require any geometrical assumption or mental reconstruction of the flow convergence and thus should be more widely applied clinically and provide better quantification of the regurgitant flow in vivo compared with two-dimensional flow convergence methods.
Few studies of three-dimensional reconstructions of flow convergence have been undertaken with color Doppler echocardiography.26 Quantitative methods derived from three-dimensionally imaged flow convergence zones have not yet been described in clinical or in vivo settings, although the importance of the three-dimensional visualization of the flow convergence zone has been appreciated.26
In the present study, using a chronic animal model with strictly quantified aortic regurgitation and a pulsatile in vitro flow model, we investigated the feasibility and value of direct measurements of three-dimensionally imaged flow convergence zones for determining the severity of aortic regurgitation.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Long-term Animal Study
For this study, we created aortic regurgitation in six juvenile sheep weighing 31 to 59 kg (mean, 42±16 kg). Six to 15 months (mean, 10±3 months) before the hemodynamic and ultrasonic studies that constitute the experimental setting for the study, the sheep had undergone thoracotomy and open heart surgery. At that time, the free edge of right coronary cusp (n=1) or the noncoronary cusp (n=5) of the aortic valve was severed with a radial incision. Subsequent aortic dilatation and/or leaflet retraction resulted in anatomic leaflet defects and failure of coaptation. All operative and animal management procedures were approved by the Animal Care and Use Committee of the NHLBI. Preoperative, intraoperative, and postoperative animal management and husbandry methods are described in detail elsewhere.27 28 During recovery, the sheep were maintained on digoxin and furosemide.
Electromagnetic Flow Probe and Meter Methods
During the experimental session, the sheep underwent repeat
thoracotomy under general anesthesia using 2% isoflurane
with oxygen. An electromagnetic flow probe (model EP455, Carolina
Medical Electronics, Inc) was placed snugly around the
pulmonary artery just above the pulmonary valve
sinuses. Another electromagnetic flow probe (model EP455, Carolina
Medical Electronics) was placed around the skeletonized ascending aorta
distal to the coronary ostia. Both flow probes were connected
to flow meters (model FM501, Carolina Medical Electronics), and these
were connected to a physiological recorder (ES
2000, Gould Inc) used for hemodynamic pressure
recordings. Aortic and left ventricular pressures
were obtained from intracavity manometer-tipped catheters (model
SPC-350, Millar Instruments, Inc) positioned transmurally. All
hemodynamic data were recorded at paper speeds of
250 mm/s. Four consecutive cardiac cycles were analyzed
for each hemodynamic determination.
The problem of the zero baseline drift was managed as previously reported20 so that the baseline for the aortic flow recording was adjusted until the forward minus the backward aortic flow volume equaled the pulmonary forward flow volume. Coronary arterial blood flow during ventricular diastole was measured in three sheep in a preliminary study. The coronary flow rate was small (0.13 to 0.23 L/min). As in other studies of aortic regurgitation, these values were considered to be negligible compared with the regurgitant volumes delineated in this study.8 The correlation coefficient for the regression of pulmonary forward flow versus aortic forward minus aortic regurgitant flow was 0.99 (SEE=0.02 L/min).
Once the curves for pulmonary and aortic flow were properly adjusted, instantaneous regurgitant flow rates could be determined, and the aortic regurgitant volumes, the integrals of instantaneous retrograde flows over diastole, were determined by planimetry of the flow signal recordings. Regurgitant fraction was calculated as retrograde aortic flow volume per minute divided by forward aortic flow volume per minute.
After baseline measurements, varying degrees of severity of aortic regurgitation were produced by altering preload and/or afterload using blood transfusion and angiotensin II (Peptide Institute Inc, provided by Tanabe Seiyaku Co). The calibrations of the flow probes were readjusted according to the manufacture's specification before each individual hemodynamic steady state, compensating for any changes in hematocrit produced by insensible fluid losses, blood loss, and/or the alteration of preload by blood transfusion. Insensible fluid loss and associated electrolyte disturbances exacerbated by the open thoracotomy were monitored by frequent (before each individual hemodynamic study) determinations of serum electrolytes and hematocrit; aberrations were avoided by continuous infusions of lactated Ringer's solution and 5% dextrose in water supplemented with potassium and calcium as necessary. Three-dimensional reconstruction study was attempted during a total of 22 steady hemodynamic states (2 to 5 per sheep).
Echocardiography and Data Acquisition
The flow convergence toward the aortic regurgitant orifice was
imaged by use of a 5-MHz transducer placed directly on the heart near
the apex with an aliasing velocity of 64 cm/s with an Interspec
ultrasound system (Apogee RX 400) (Fig 1
). Guided by the two-dimensional and
color Doppler imaging of the regurgitant jet and the valve, CW
Doppler recordings of the regurgitant flow velocity
parallel to the direction of the aortic regurgitant jet were performed.
This CW velocity profile was integrated over time to determine
regurgitant volumes per beat.
|
In our previous in vitro steady flow studies,29 we had
observed the influences of instrument settings for color Doppler
flow mapping on transferring the color flow mapping data into a
black-and-white video composite data milieu for three-dimensional
reconstruction. After trying several different color Doppler flow
mapping settings for three-dimensional reconstruction, we determined
that a red to yellow to blue aliased velocity and nonvariance color
encoding produced the most clearly defined surface zones of color
Doppler flow convergence (Fig 1
). The
echocardiographic probe was mounted on a holding gantry
that positioned the probe on the apex of the heart in a prototype
stepper motor system that was controlled by a dedicated
three-dimensional image processing computer (TomTec Imaging System).
The stepper motor, which was driven by a steering logic in the TomTec
computer, allowed rotation of the probe at any desired increment
between 0° to 180° while the probe was scanning the heart. With
1° increments of probe rotation, 180 slices of the flow convergence
region were obtained over the entire scan arc (180°) for each
hemodynamic condition and transferred during
acquisition into the TomTec computer as previously
reported.26 The scanning and acquisition of the color
Doppler flow mapping data were gated to the ECG at heart rates of
91 to 136 bpm and to the respiratory cycles. An ECG gating interval of
<20% of the RR interval (less than ±40 ms) and respiratory gating
within limits between inspiratory and expiratory phase were
predetermined before image acquisition using the "observe"
function of the instrument. When the ECG and respiratory gating met the
predetermined limits, video composite images were acquired at each
33-ms interval (30 frames per second) after the R-wave signal. Time
resolution (frame rate) of the three-dimensionally reconstructed images
was not limited by the TomTec system or the three-dimensional method
but was, in fact, limited to 12 to 17 frames per second, which were the
original color Doppler acquisition frame rates. Image acquisition
took a mean of 112±56 seconds to accomplish. Once the scanning
sequence was completed, the digital images were stored for
post-processing.
In Vitro Study
To study central regurgitant flow and to minimize the effects of
flow confinement in determining accelerating flow geometry proximal to
the regurgitant orifice, we also performed an in vitro pulsatile flow
study. Because the mean aortic diameter in the animal study was 1.8 cm
and thus may have created more constraint of the convergence flow than
might be clinically encountered in an enlarged aortic root, we designed
a model with a plastic cylinder with a diameter of 3.3 cm around a
regurgitant orifice, mimicking the size of an adult aorta. We created a
circular central orifice (area=0.24 cm2) in a plastic disk
that had three leaflets, mimicking an aortic regurgitant orifice not
closely bounded by the aortic wall or sinuses. In this flow model,
which has been described previously,29 pulsatile flows
were generated into the inlet chamber using a pulsatile pump (Harvard
piston pump, model 1423). Flow passed through the regurgitant orifice
into the outlet chamber and returned to the reservoir.29
Because the maximal regurgitant flow rate was 5.6 L/min in the animal
study, a wider range of peak regurgitant flow rates (from 3.2 to 15.2
L/min) was generated in this in vitro pulsatile flow study,
investigating the applicability of the three-dimensional method for
greater ranges of severity of aortic regurgitation.
Actual regurgitant flow rates were obtained by a ultrasonic flow probe
(model 16NB272, Transonic Systems Inc) and meter (model T106X,
Transonic Systems Inc) that were connected to the flow model. We used
the same ultrasound system with a 50-MHz transducer and the TomTec
system as the animal study. An aliasing velocity of 28 cm/s was
selected for imaging flow convergence regions. An electromagnetic
device was attached to this pulsatile pump to provide an "ECG"
signal into the TomTec system for gating the acquisition and the
transfer of the color Doppler two-dimensional images.
Three-dimensional Reconstruction
After image alignment, a process of feature extraction and
interpolation by the TomTec computer filled in the gaps between slices
to obtain the reconstruction and surface rendering of the flow
convergence zones. For most hemodynamic conditions,
only the aliased boundaries were transferred into the TomTec system for
three-dimensional reconstruction. This was accomplished by use of
relatively low color Doppler gains and high wall filters to
emphasize the brightness of the boundary of the alias. When this could
not be satisfactorily accomplished, unaliased accelerating flow
proximal to the flow convergence boundary as well as the aliased flow
convergence boundary were transferred, as shown in Fig 1. However,
under these conditions, the brightest boundary surface that
corresponded to the aliasing velocity could be selected later using the
TomTec's "threshold" function before reconstruction so that the
flow convergence surface could be reliably reconstructed. The resulting
three-dimensionally reconstructed images could be inspected from any
desired viewing perspectives, although the aliasing velocity could not
be changed in the TomTec system (Fig 2).
Gray scale, surface rendering, and image resolution provided by the
TomTec computer were optimized to obtain clearly defined isovelocity
surfaces, which were recognized as the brightest demarcated surfaces
proximal to the aortic regurgitant orifices. After determining the
exact spatial locations and temporal changes of the flow convergence
zones, we selected and magnified the maximal flow convergence zone,
which was usually observed in early to mid diastole, for
later analyses. The timing for measuring the maximal flow
convergence was also selected using the ECG as well as the flow image
on the monitor screen of the TomTec system.
|
Estimation of the Severity of Aortic Regurgitation
Direct Measurement of the Maximal Surface Area of the Flow
Convergence Zones Without Geometric Assumption in the Chronic
Animal Study
The maximal flow convergence zones were cut in parallel sections
at 0.3- to 1.0-mm intervals using the software of the TomTec computer
(Fig 3). By use of the computer
trackball, the arc length of the flow convergence boundary in each
sectioned plane (bold boundary arc lines in the right of Fig 3) was
measured as shown in Fig 3. The boundary arc lengths of each parallel
slice of the flow convergence were multiplied by the slice thickness;
then these values were sequentially added, resulting in the entire
surface area of the maximal flow convergence zone (S). Peak aortic
regurgitant flow rates were calculated based on the continuity concept
(Q=S · V, where V is aliasing velocity).
|
Conventional Two-dimensional Flow Convergence Methods
To compare the three-dimensional method with the conventional
two-dimensional color Doppler flow convergence methods for
evaluating the severity of the aortic regurgitation,
two-dimensional color Doppler images recorded on super VHS
videotapes during probe rotation also were analyzed. From these
videotape records, we selected and measured the maximal axial
distance from the orifice to the clearly imaged isovelocity surface (r)
at the same aliasing velocity (V) as was used for three-dimensional
imaging. The regurgitant orifice position was defined as the smallest
connection between the flow convergence and the regurgitant jet. Peak
regurgitant flow rates were calculated by use of a simple hemispherical
model (Q=2
r2 · V).
We also measured three orthogonal axes of the flow convergence region
in two orthogonal planes (a=major axis, b=minor axis, and c=height)
using the videotapes of the two-dimensional images obtained during the
probe rotation for the three-dimensional data acquisition. Using the
following equation,13 we calculated the maximal surface
area of the flow convergence region:
 |
 |
 |
Calculated peak regurgitant flow rates using the three-dimensional and two-dimensional methods were compared with electromagnetically obtained peak regurgitant flow rates. Among the indexes of aortic regurgitant severity, only the peak regurgitant flow rate was directly related to the maximal flow convergence surface area. Because one of our previous studies revealed little dynamic change in aortic regurgitant orifice area during diastole,30 regurgitant volume per beat was also obtained noninvasively as the product of the calculated peak regurgitant flow rate and the ratio of the velocity-time integral to the CW peak velocity through the regurgitant orifice; these values were compared with those obtained from the electromagnetic flow meters.
In Vitro Study
After the reconstruction of the flow convergence zones (Fig 4), direct measurements of the maximal
flow convergence zones were performed as described for the animal
study. Two-dimensional methods using the hemispherical and
hemielliptical models were also performed as in the animal study to
estimate the maximal flow convergence surface areas. Then peak flow
rates were calculated as the products of the surface areas and the
aliasing velocities used. Three-dimensionally and two-dimensionally
calculated peak flow rates were compared with those obtained by the
ultrasonic flow meter.
|
Interobserver Variability
To evaluate the effect of observational variability on the
measurement of volumes of the maximal flow convergence zones and peak
regurgitant flow rates, 10 randomly selected flow conditions were
analyzed at different times with the same computer by two
independent observers, each without knowledge of the results obtained
by the other or the actual flow data.
Statistical Analysis
Data are presented as mean±SD for descriptive
statistics. Because multiple points were used in the same sheep in the
animal study, the relationship between the two-dimensional methods and
the three-dimensional method of flow determination versus the
electromagnetic flow meter method was analyzed by use of a
univariate ANCOVA. One ANCOVA model was fitted for each of
the three calculation methods versus the electromagnetic flow meter
method to give the estimates of the regression coefficients (ie, the
intercepts and slopes). Each of these ANCOVAs included sheep as a
factor with dummy variable coding to control for sheep to sheep
differences.31 The estimates for these ANCOVAs were
presented as follows: y=(common
slope)x+(averaged intercept for all sheep), where
y is the two- or three-dimensionally obtained value,
x is the value obtained by the electromagnetic flow meters,
and common slope refers to the slope of the regression common to all
sheep. To compare the two-dimensional methods with the
three-dimensional method, a combined ANCOVA model that included both
the method of flow determination and sheep (again with dummy
variable coding) as factors was used. Because the interest here is
to compare the performance of the two-dimensional methods with
the three-dimensional method, the comparisons of intercepts and slopes
in this ANCOVA were as follows: the two-dimensional hemispherical
method versus the three-dimensional method and the two-dimensional
hemielliptical method versus the three-dimensional method (ie, these
were the contrasts used in the model).
To assess agreement and predictability between the actual flow rates and three-dimensional and two-dimensional image–based calculated flow rates, the method of Bland and Altman32 was used. Statistical analyses were performed by use of the statistical package S-PLUS (S-PLUS for Windows, version 3.2 supplement, StatSci Division of Mathsoft, Inc, 1994). Statistical significance was defined as a value of P<.05.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Long-term Animal Study
For all 22 hemodynamic conditions except 2, video composite data of color Doppler flow mapping images of the flow convergence were transferred successfully into the TomTec computer, and diastolic frames with the largest flow convergence zones could be selected from the reconstructed three-dimensional image loops. The mean reconstruction time was 15 minutes for highest-resolution three-dimensional images, including the time for the initial low-resolution reconstructions.
Severity of Aortic Regurgitation
Aortic regurgitant volumes and regurgitant fractions for the
remaining 20 hemodynamic conditions were within
clinically relevant ranges of mild to moderate aortic
regurgitation from 1.0 mL/beat to 23 mL/beat (average
11±6.3 mL/beat) and from 3% to 42% (average 24±12%), respectively.
Peak and mean regurgitant flow rates were also within clinically
relevant ranges from 1.2 to 5.6 L/min (average 2.9±1.4 L/min) and from
0.1 to 2.3 L/min (average 1.1±0.7 L/min), respectively.
Evaluation of the Severity of Aortic
Regurgitation
There were better agreements between the three-dimensionally
calculated peak flow rates and electromagnetically derived peak
regurgitant flow rates than those between the two-dimensionally
calculated peak flow rates and the reference values (r=.94,
difference=0.02±0.29 L/min for the three-dimensional method;
r=.80, difference=1.2±1.7 L/min for the two-dimensional
hemispherical model; r=.75, difference=-0.20±0.94 L/min
for the two-dimensional hemielliptical model, Fig 5). ANCOVA was used to eliminate the
effect of multiple points from the same sheep and showed that peak
regurgitant flow rates obtained from the three-dimensional method best
agreed with those obtained electromagnetically
(y=1.0x-0.16, Fig 5A1). It revealed a
significant difference in the slope of the regression between the
two-dimensional hemispherical method (y=1.6x
-0.3, P<.001; Fig 5B1) and the two-dimensional
hemielliptical method (y=0.90x+ 0.2,
P<.002; Fig 5C1) versus the three-dimensional method
(probability values are from comparisons to the slope of the regression
between the three-dimensional and the electromagnetically derived peak
regurgitant flow rates). This showed that the two-dimensional methods
differ from the three-dimensional method (with an estimated slope of 1)
and are significantly different from a reference identity (a line with
a theoretical slope of 1).
|
Regurgitant volumes calculated by the three-dimensional method agreed better with those electromagnetically obtained measurements than did the two-dimensional hemispherical and hemielliptical methods (r=.92, y=0.98x -0.3, difference=1.2±2.4 mL per beat for the three-dimensional method; r=.81, y=1.6x-0.7 [P<.001], difference=7.7±7.8 mL per beat for the hemispherical model; r=.68, y=0.92x+2.3 [P<.05], difference=-3.2±6.5 mL per beat for the hemielliptical model) (both as compared to the slope of the regression between the three-dimensional and the electromagnetically derived regurgitant volumes).
The conventional hemispherical two-dimensional method substantially overestimated the peak regurgitant flow rates and regurgitant volumes.
In Vitro Study
As Fig 4 shows, skewed hemielliptical shapes of the flow
convergence zones were seen on the three-dimensional reconstruction
images. Fig 6 shows an excellent
agreement between the three-dimensionally obtained peak flow rates and
those obtained by the flow meter (r=.99,
P<.0001, y=0.98x-0.2, SEE=0.06
L/min, mean difference=-0.16±0.20 L/min). As observed in the animal
study, both two-dimensional methods provided less agreement with
reference peak regurgitant flow rates and greater data scatters
(r=.94, P<.001,
y=1.38x-1.6, SEE=1.4 L/min, mean
difference=0.75±1.8 L/min for the hemispherical model;
r=.90, P<.002, y=0.93x+
0.12, SEE=1.7 L/min, mean difference=-0.77±1.6 for the hemielliptical
model).
|
Interobserver Variability
There was a good agreement between the two independent observers'
measurements for three-dimensionally calculated peak regurgitant flow
rates for the animal study and the in vitro study (r=.89,
mean difference=0.26±0.24 L/min; r=.96, mean
difference=0.13±0.15 L/min, respectively).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Previous Flow Convergence Studies
The flow convergence phenomenon has been used for quantifying aortic regurgitation, as well as mitral regurgitation, in in vitro and in vivo studies.11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 30 Previous reports attempting to quantify regurgitant flow rates and flow orifice areas with the flow convergence phenomenon assumed a simple or modified hemispherical or hemielliptical shape of the isovelocity surface using two-dimensional imaging of the flow convergence.13 14 15 16 17 18 19 20 21 22 23 24 25 Based on the continuity concept, the regurgitant or forward flow rates in these articles were calculated as the product of the isovelocity surface area and the aliasing velocity used for imaging the flow convergence. Thus, accurate measurements of the isovelocity surface area are of primary importance for correctly estimating the flow rate. The geometry of the flow convergence isovelocity surface, however, depends on many factors, including regurgitant flow rate, the selected aliasing velocity, orifice geometry, surrounding, confining structures, and pathophysiological factors such as left ventricular compliance.13 15 16 25 Even when the isovelocity shape appears to be a hemisphere in one plane in a typical view, such as an apical four chamber view, it may not actually be a hemisphere. The converse may also be true. Because of both the complex acceleration field geometry and because of the angle dependence of the color Doppler flow imaging, in our previously performed two-dimensional flow convergence studies, we encountered difficulties in judging the geometry of the flow convergence isovelocity surface, often resulting in underestimation or overestimation of the electromagnetically obtained regurgitant flow rates.30 33 However, in these experimental studies, the best aliasing velocity for the simple hemispherical assumption could be selected to determine the regurgitant flow rates to match those from the electromagnetic flow method after the direct transfer of the digital flow convergence velocity assignment into the microcomputer or by use of complicated selection formulas based on regurgitant flow rates and velocities.30 33 These methods often are not practical in clinical settings.
Advantages of the Three-dimensional Flow Convergence
Method
With the three-dimensional method, one does not need to assume any
specific geometric configuration such as that of a hemisphere.
Variations in confining structures surrounding aortic regurgitant
orifices in patients and selected aliasing velocity can create
important variations in the shape of, and skew of, the flow convergence
and thus in the flow measurements. The three-dimensional method should
be less prone to error than two-dimensional methods because the surface
area of any flow convergence geometry can be determined by the
three-dimensional method used in the present study, eliminating the
need for imaginative mental reconstruction or assumptions regarding its
three-dimensional geometry from two-dimensional observations.
Additionally, one can use a wider range of aliasing velocities
permitting the flow convergence surface to be clearly imaged. This
advantage may be of clinical importance since cumbersome efforts for
selecting the optimal aliasing velocity for assuming a certain geometry
of a flow convergence surface, such as a hemisphere, can be
avoided.
Study Limitations
In this study, the echo-Doppler imaging was performed with the
transducer directly on the heart, somewhat reducing translational heart
motion problems. Thus, the quality of the original two-dimensional
color Doppler imaging may have been better than that obtained in
clinical settings.30 33 However, because the
three-dimensional data sets were derived from the two-dimensional
imaging during the probe rotation, some of the heart motion artifacts
may have degraded the three-dimensional reconstruction images. The
comparison between two- and three-dimensional images is thus somewhat
biased in favor of the two-dimensional data analyses by this
limitation.
Limitations inherent to the color Doppler flow mapping for imaging the flow convergence, including instrument factors such as color gain, wall filter settings, and variability of aliasing velocities, are carried into the three-dimensionally reconstructed flow convergence images. Low frame rates (12 to 17 frames per second) may cause underestimation of the maximal flow convergence size. However, this underestimation resulting from frame selection should be minimal because relatively constant aortic regurgitant flow and orifice area throughout diastole have been observed in our earlier studies.20 30
Especially important is the loss of velocity information for flows at the edges of the flow convergence region induced by the angle between Doppler interrogation and the actual direction of blood flow. Because of these problems, portions of the imaged flow convergence surface adjacent to the valves in this study do not correspond strictly to true isovelocity surfaces. Thus, this technique needs correction for flow constraint, competing flows, and Doppler angle dependency. Imaging from more than one site or compound image rasters recently reported for real-time volume imaging would be desirable.34 35 Angle-independent methods, such as magnetic resonance imaging and digital particle tracking, should also provide insight into methods for correcting angle dependent color Doppler data.16
In a recent in vitro study,29 probably because of the Doppler angle problems discussed above, we encountered significant underestimation using the three orthogonal measurements for three-dimensional reconstruction of the flow convergence zones used to calculate steady flow rate through a restrictive orifice. However, in the present study, no such underestimation was observed in either the in vivo or in vitro study. Several factors may have been involved. Higher pulsatile regurgitant flows with the maximal flow rate of 5.6 L/min observed in the present in vivo study and 15.2 L/min in the in vitro pulsatile flow study as opposed to lower steady flow rates with the maximal flow rate of 2.4 L/min in the in vitro steady flow study may have lessened the underestimation. However, the range of the severity of regurgitation examined in the present animal and in vitro pulsatile flow studies was closer to clinical conditions in terms of volume and flow rate than that in the previously performed steady flow study. Postmortem inspection of the aortas and three-dimensional reconstruction of the anatomic configuration of the aortic valve leaflets showed complicated, curved structures surrounding the regurgitant orifice, which may have caused confinement of convergent flow. As opposed to the flat orifice used in the previous in vitro steady flow study,29 the geometrically complicated, nonflat orifice and constraint of flow observed near the walls of the aorta in the present in vivo study would decrease the amount of flow perpendicular to and increase the flow parallel to the Doppler interrogation, lessening the Doppler angle effect. This effect may be exaggerated in the animal study because of small aortas (mean diameter=1.8 cm) compared with those of adult humans. Thus, in the in vitro pulsatile flow study, the simulated aorta was made larger than the animal aorta so that clinically relevant flow constraint could be studied. Not only in the animal study but also in the in vitro pulsatile study, very good agreement was present between values obtained by the reference flow meter methods and those by the three-dimensional method. The present three-dimensional method therefore appears to be valid for evaluating clinically encountered aortic regurgitation. Although probably of lesser importance, certain technical issues related to the data acquisition need to be considered. High color Doppler wall filters used for obtaining distinct isovelocity contours may result in overestimation. The center axis of probe rotation may change during data acquisition. In addition, averaging data-fill algorithms for voxels may cause deterioration of lateral resolution and expand the spatial extent of the flow data set, resulting in overestimation of the lengths of the axes which are then multiplied.
The relatively high cost of the additional equipment, the time required to process the images, and the limitations of the measurement algorithms supported may hinder immediate clinical application of three-dimensional flow computation methods. In the future, however, continuing development of computer technology and ultrasound equipment should provide direct transfer of color-encoded signals for three-dimensional reconstruction of color Doppler flow images derived as real-time volume images to improve further visualization and quantification of cardiovascular flow phenomena.
Conclusions
In our study, direct measurements of three-dimensionally
reconstructed proximal isovelocity flow convergence surface areas
provided more accurate regurgitant flows than conventional
two-dimensional color Doppler methods.
|
Acknowledgments |
|---|
This study was supported in part by a grant (HL-43287) from the National Heart, Lung, and Blood Institute. We acknowledge the assistance of the veterinary professional and technical staff of the Laboratory of Animal Medicine and Surgery, NHLBI, and the dedication of Michelle Ann Cheney for help with editing and preparing the manuscript.
|
Footnotes |
|---|
Reprint requests to Dr Michael Jones, National Institutes of Health, NHLBI, 9000 Rockville Pike, Bldg 14E, Room 107A, Bethesda, MD 20892.
Received October 3, 1996; revision received June 17, 1997; accepted June 26, 1997.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1. Veyrat C, Lessana A, Abitbol G, Ameur A, Benaim R, Kalmanson D. New indexes for assessing aortic regurgitation with two-dimensional Doppler echocardiographic measurement of the regurgitant aortic valvular area. Circulation. 1983;68:998-1005.
2. Omoto R, Yokote Y, Takamoto M, Kyo S, Ueda K, Asano H, Namekawa K, Kasai C, Kondo Y, Koyano A. The development of real-time two-dimensional Doppler echocardiography and its clinical significance in acquired valvular diseases. Jpn Heart J. 1984;25:325-340.[Medline] [Order article via Infotrieve]
3. Miyatake K, Okamoto M, Kinoshita N, Izumi S, Owa M, Takao S, Sakakibara H, Nimura Y. Clinical application of a new type of real-time two-dimensional Doppler flow imaging system. Am J Cardiol. 1984;54:857-868.[Medline] [Order article via Infotrieve]
4. Perry GJ, Helmcke F, Nanda NC, Byard C, Soto B. Evaluation of aortic insufficiency by Doppler color flow mapping. J Am Coll Cardiol. 1987;9:952-959.[Abstract]
5. Grayburn PA, Handshoe R, Smith MD, Harrison MR, DeMaria AN. Quantitative assessment of the hemodynamic consequences of aortic regurgitation by means of continuous wave Doppler recordings. J Am Coll Cardiol. 1987;10:135-141.[Abstract]
6. Bouchard A, Yock P, Schiller NB, Blumlein S, Botvinick EH, Greenberg B, Cheitlin M, Massie BM. Value of color Doppler estimation of regurgitant volume in patients with chronic aortic insufficiency. Am Heart J. 1989;117:1099-1105.[Medline] [Order article via Infotrieve]
7. Smith, MD, Kwan OL, Spain MG, DeMaria AN. Temporal variability of color Doppler jet areas in patients with mitral and aortic regurgitation. Am Heart J. 1992;123:953-960.[Medline] [Order article via Infotrieve]
8. Reimold SC, Ganz P, Bittl JA. Effective aortic regurgitant orifice area: description of a method based on the conservation of mass. J Am Coll Cardiol. 1991;18:761-768.[Abstract]
9. Holm S, Eriksson P, Karpo K, Osterman G, Teien D. Quantitative assessment of aortic regurgitation by combined two-dimensional, continuous-wave and color flow Doppler measurements. J Int Med. 1992;231:115-121.[Medline] [Order article via Infotrieve]
10.
Sutton DC, Kluger R, Ahmed SU, Reimold SC, Mark
JB. Flow reversal in the descending aorta: a guide to
intraoperative assessment of aortic regurgitation with
transesophageal
echocardiography. J Thorac
Cardiovas Surg. 1994;108:576-582.
11.
Yoshida K, Yoshikawa J, Hozumi T, Akasaka T, Yamaura Y,
Minagoe S, Nakanishi H, Miyamoto T. Assessment of aortic
regurgitation by the acceleration flow void proximal to
the leaking orifice in cine magnetic resonance imaging.
Circulation. 1991;83:1951-1955.
12. Cranney GB, Benjelloun H, Perry GJ, Lotan CS, Blackwell JG, Coghlan CH, Pohost GM. Rapid assessment of aortic regurgitation and left ventricular function using cine nuclear magnetic resonance imaging and the proximal convergence zone. Am J Cardiol. 1993;71:1074-1081.[Medline] [Order article via Infotrieve]
13. Utsunomiya T, Ogawa T, Doshi R, Patel D, Quan M, Henry WL, Gardin JM. Doppler color flow `proximal isovelocity surface area' method for estimating volume flow rate: effects of orifice shape and machine factors. J Am Coll Cardiol. 1991;17:1103-1111.[Abstract]
14.
Recusani F, Bargiggia GS, Yoganathan AP, Raisaro A,
Valdes-Cruz LM, Sung HW, Bertucci C, Gallati M, Moises VA, Simpson IA,
Tronconi L, Sahn DJ. A new method for quantification of
regurgitant flow rate using color Doppler flow imaging of the flow
convergence region proximal to a discrete orifice: an in vitro
study. Circulation. 1991;83:594-604.
15.
Rodriguez L, Anconina J, Flachskampf FA, Weyman AE,
Levine RA, Thomas JD. Impact of finite orifice size on
proximal flow convergence: implication for Doppler quantification
of valvular regurgitation. Circ
Res. 1992;70:923-930.
16. Shandas R, Gharib M, Liepmann D, Shiota T, Sahn DJ. Experimental studies to define the geometry of the flow convergence region: Laser Doppler particle tracking and color Doppler imaging. Echocardiography. 1992;9:43-50.[Medline] [Order article via Infotrieve]
17. Zhang J, Jones M, Shandas R, Valdes-Cruz LM, Murillo A, Yamada I, Kang SU, Weintraub RG, Shiota T, Sahn DJ. Accuracy of flow convergence estimates of mitral regurgitant rates obtained by using multiple color flow Doppler M-mode aliasing boundaries: an experimental animal study. Am Heart J. 1993;22:1522-1592.
18.
Deng YB, Shiota T, Shandas R, Zhang J, Sahn DJ.
Determination of the most appropriate velocity threshold for applying
hemispherical flow convergence equation to calculate flow rate
according to the transorifice pressure gradient: Digital computer
analysis of Doppler color flow convergence region.
Circulation. 1993;88:1699-1708.
19. Shiota T, Teien D, Deng YB, Shandas R, Sahn DJ. Estimation of regurgitant flow volume on centerline velocity profiles using digital color M-Q Doppler applied to orifices of different shapes. J Am Coll Cardiol. 1994;24:440-445.[Abstract]
20. Shiota T, Jones M, Yamada I, Heinrich RS, Ishii M, Sinclair B, Yoganathan AP, Sahn DJ. Evaluation of aortic regurgitation using digitally determined color Doppler-imaged flow convergence acceleration: a quantitative study in animals. J Am Coll Cardiol. 1996;27:203-210.[Abstract]
21.
Rodriguez L, Thomas JD, Monterroso V, Mueller L,
Harrigan H, Weyman RA. Validation of the proximal convergence
method: calculation of orifice area in patients with mitral
stenosis. Circulation. 1993;88:1157-1165.
22. Deng YB, Matsumoto M, Wang XF, Takizawa S, Takekoshi N, Shimizu T, Mishima K. Estimation of mitral valve area in patients with mitral stenosis by the flow convergence region method: selection of aliasing velocity. J Am Coll Cardiol. 1994;24:683-689.[Abstract]
23.
Min Pu, Vandervoort PM, Griffin BP, Leung DY, Stewart
WJ, Cosgrove DM III, Thomas JD. Quantification of mitral
regurgitation by the proximal convergence method using
transesophageal echocardiography:
clinical validation of a geometric correction for proximal flow
constant. Circulation. 1995;92:2169-2177.
24. Cape EG, Thomas JD, Weyman AE, Yoganathan AP, Levine RA. Three-dimensional surface geometry correction is required for calculating flow by the proximal isovelocity surface area technique. J Am Soc Echocardiogr. 1995;8:585-594.[Medline] [Order article via Infotrieve]
25. Schwammenthal E, Chen C, Giesler M, Sagie A, Guerro JL, Vazquez de Prada JA, Hombach V, Weyman AE, Levine RA. New method for accurate calculation of regurgitant flow rate based on analysis of Doppler color flow maps of the proximal flow field: validation in a canine model of mitral regurgitation with initial application in patients. J Am Coll Cardiol. 1996;27:161-172.[Abstract]
26. Delabays A, Sugeng L, Pandian NG, Hsu TL, Ho SJ, Chen CH, Marx G, Schwartz S, Cao QL. Dynamic three-dimensional echocardiographic assessment of intracardiac blood flow jets. Am J Cardiol. 1995;76:1053-1058.[Medline] [Order article via Infotrieve]
27. Jones M, Barnhart GR, Chavez AM, Jett GK, Rose DM, Ishihara T, Ferrans VJ. Experimental evaluation of bioprosthetic valves implanted in sheep. In: Cohn LM, Gallucci V, eds. Cardiac Bioprostheses. New York, NY: York Medical Books; 1982:275-292,.
28.
Tamura K, Jones M, Yamada I, Ferrans VJ. A
comparison of failure modes of glutaraldehyde-treated
versus antibiotic-preserved mitral valve allografts implanted in
sheep. J Thorac Cardiovasc Surg. 1995;110:224-238.
29. Shiota T, Sinclair B, Ishii M, Zhou X, Ge S, Teien DE, Gharib M, Sahn DJ. Three-dimensional reconstruction of color Doppler flow convergence regions and regurgitant jets: an in vitro quantitative study. J Am Coll Cardiol. 1996;27:1511-1518.[Abstract]
30.
Shiota T, Jones M, Yamada I, Heinrich RS, Ishii M,
Sinclair B, Holcomb SR, Yoganathan AP, Sahn DJ. Effective
regurgitant orifice area by the color Doppler flow convergence
method for evaluating the severity of chronic aortic
regurgitation: an animal study.
Circulation. 1996;93:594-602.
31. Glantz SA, Slinker BK. Repeated measures. In: Primer of Applied Regression and Analysis of Variance. New York, NY; McGraw-Hill Book Co; 1990:381-463.
32. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurements. Lancet. 1986;1:307-310.[Medline] [Order article via Infotrieve]
33. Shiota T, Jones M, Teien DE, Yamada I, Passafini A, Ge S, Sahn DJ. Dynamic change in mitral regurgitant orifice area: comparison of color Doppler echocardiographic and electromagnetic flow meter-based methods in a chronic animal model. J Am Coll Cardiol. 1995;26:528-536.[Abstract]
34.
Mizushige K, Shiota T, Paik J, Griever G, Depp MR,
Passafini A, Shandas R, DeMaria AN, Sahn DJ. Effects of
pulmonary venous flow direction on mitral regurgitant jet area
as imaged by color Doppler flow mapping: an in vitro study.
Circulation. 1995;91:1834-1839.
35. Ota T, Fleishman CE, Li J, Stetten G, Ohazama C, von Ramm O, Kisslo J. Real-time three dimensional echocardiography with saline contrast enhancement: methods and possibilities. J Am Coll Cardiol. 1996;27:405A. Abstract.
This article has been cited by other articles:
 |
|
 |
|
  T Irvine, X K Li, D J Sahn, and A Kenny Assessment of mitral regurgitation Heart, November 1, 2002; 88(90004): iv11 - 19. [Full Text] [PDF]
|
|
|
|
 |
|
 M. Ishii, K. Hashino, G. Eto, T. Tsutsumi, W. Himeno, Y. Sugahara, H. Muta, J. Furui, T. Akagi, Y. Ito, et al. Quantitative Assessment of Severity of Ventricular Septal Defect by Three-Dimensional Reconstruction of Color Doppler-Imaged Vena Contracta and Flow Convergence Region Circulation, February 6, 2001; 103(5): 664 - 669. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T Irvine, X N Li, R Rusk, D Lennon, D J Sahn, and A Kenny Three dimensional colour Doppler echocardiography for the characterisation and quantification of cardiac flow events Heart, November 1, 2000; 84(90002): 2i - 6. [Full Text]
|
|
|
|
|
|
Y. Mori, T. Shiota, M. Jones, S. Wanitkun, T. Irvine, X. Li, A. Delabays, N. G. Pandian, and D. J. Sahn Three-Dimensional Reconstruction of the Color Doppler–Imaged Vena Contracta for Quantifying Aortic Regurgitation : Studies in a Chronic Animal Model Circulation, March 30, 1999; 99(12): 1611 - 1617. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||