From the Departments of Cardiology (J.Y., A.D., J.R.T.C.R.), Surgery
(M.R.H.M. v S.), Radiology (L.C. v D.), and Neurology (P.J.K.), Thoraxcenter
and Academic Hospital Dijkzigt, Erasmus University, Rotterdam, the
Netherlands, and Esaote Biomedica, SpA (M.K.), Florence, Italy.
Correspondence to Prof Dr Jos R.T.C. Roelandt, Room Bd 408, Thoraxcenter, 3015 GD Rotterdam, Netherlands. E-mail roelandt{at}card.azr.nl
Methods—We studied 14 patients with both carotid angiography and
3-D ultrasound. Of 13 patients who underwent surgery, 12 were
reexamined with 3-D ultrasound after surgery. The length and volume of
20 randomly selected plaques were measured from 3-D data sets. The
severity of stenosis was quantified by 3-D ultrasound using
both a diameter method and an area method on cross-sectional views at
the most stenotic site; the results were then compared with
those from carotid angiography. The segmental vessel volume and average
cross-sectional area of the operated artery both before and after
endarterectomy were measured from 3-D ultrasound
data.
Results—Good correlation was obtained between 3-D ultrasound and
carotid angiography in quantitative analysis of carotid
stenosis (SEE=12.4%, r=0.76, and mean
difference=7.0±12.3% with the diameter method; SEE=10.5%,
r=0.82, and mean difference=1.8±10.5% with the area
method by 3-D ultrasound). 3-D ultrasound had excellent reproducibility
and small intraobserver and interobserver variability in plaque length
and volume measurements. No significant changes in segmental vessel
volume and average cross-sectional area of the operated artery were
observed after surgery in patients with suture closure. However, a
significant increase in segmental vessel volume was obtained in
patients with polyfluorethylene patches applied to the surgical opening
of the artery.
Conclusions—3-D ultrasound can be used for both qualitative and
quantitative analysis of plaques in the carotid artery and to
detect and quantify significant carotid stenosis. Its
volumetric potential has important clinical implications in serial
follow-up studies for observing the progression or regression of
stenotic lesions and for evaluating the outcome of
interventional procedures such as endarterectomy or
stent placement.
Carotid Artery Angiography
Carotid Endarterectomy
3-D Ultrasound of Carotid Arteries
Data Acquisition and Processing
Data Analysis
By moving the cutting plane through the 3-D data set in various
directions, multiple views of the carotid artery were produced. The
length of a plaque was measured in a longitudinal view of the vessel.
By defining both ends of the plaque, multiple (as many as 20)
equidistant cross-sectional cutting planes perpendicular to the
longitudinal view of the plaque were generated automatically. The
volume of the plaque was computed automatically from manual tracing of
its borders on each cross-sectional image using a "summation of
discs" method. The severity of stenosis was computed from
measurements on the cross-sectional image with the smallest free lumen
area with the following 2 methods. With the diameter method, the
severity of stenosis was derived from measurements of the
smallest free lumen diameter and the diameter of the original vessel
lumen. With the area method, it was derived from the cross-sectional
free lumen area and the original vessel lumen area (Figure 2
In patients who underwent carotid endarterectomy,
segmental vessel volume of the carotid artery bifurcation was measured
before and after surgery. A longitudinal cutting plane of the carotid
artery bifurcation showing the common, internal, and external carotid
arteries was selected as a reference image from the 3-D data set. A
20-mm segment of the vessel, including 10 mm of the common carotid
artery proximal and 10 mm of both internal and external carotid
arteries distal from the bifurcation, was defined. Twenty parallel
equidistant cross-sectional images of this segment were automatically
generated, with each slice 1 mm in thickness. The area of the
carotid artery within the adventitia, ignoring any plaques if
present, was traced manually on each slice to derive the segmental
vessel volume. Divided by the predefined length of the segment (20
mm), an average cross-sectional area of the carotid bifurcation was
also computed.
Statistics
Carotid Artery Angiography
3-D Ultrasound of the Carotid Artery
Carotid Stenosis
Plaques
Surgical Impact
Additional Information
3-D Ultrasound of the Carotid Artery
Changes of the segmental vessel volume and, therefore, the
average cross-sectional area of the original carotid artery (ignoring
the plaques) increased after endarterectomy in
patients with patches used on closing of the artery, as can be
expected. In patients without application of patches, changes in
segmental vessel volumes after surgery were less significant and varied
from decrement to increment. This is an interesting observation,
although the answer to it is not clear. We believe that the change in
segmental vessel volume of the carotid artery after surgery is
multifactor dependent. On one hand, the suture might decrease the size
of the vessel, resulting in decrease of segmental vessel volume and
cross-sectional area. On the other hand, removal of the intima-media
complex and the plaque (especially the calcified ones) might increase
the distensibility and decrease the recoil force of the involved
segment of the vessel, resulting in increase of both segmental volume
and cross-sectional area. Better understanding of the impact of
endarterectomy on the carotid artery requires
further investigation with a larger number of patients.
Limitations of the Study
Conclusions
Received March 2, 1998;
revision received June 5, 1998;
accepted July 27, 1998.
© 1998 American Heart Association, Inc.
Original Contributions
Three-Dimensional Ultrasound Study of Carotid Arteries Before and After Endarterectomy
Analysis of Stenotic Lesions and Surgical Impact on the Vessel
Abstract
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Background and Purpose—It has
been proved that symptomatic patients with severe carotid
stenosis benefit from endarterectomy.
Currently used methods for quantitation of the severity of carotid
stenosis have limitations, and the impact of
endarterectomy on the operated region of carotid
artery remains unknown. The purpose of this study was to examine the
accuracy of a 3-D ultrasound system for quantitation of
stenotic lesions and to evaluate changes in regional vessel
volume and cross-sectional area after carotid endarterectomy.
Key Words: atherosclerosis • carotid arteries • carotid endarterectomy • carotid stenosis • ultrasonography
Introduction
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
It has been proved that symptomatic
patients with severe carotid artery stenosis (
70%) benefit
from endarterectomy.1 2
Selection of candidates for surgery is based mainly on carotid
angiography.3 4 5 6 Because of the invasiveness and
the associated complications and mortality of this technique,
noninvasive procedures have been used for evaluating the severity of
carotid artery stenosis.7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 2-D
ultrasound scanning, in combination with Doppler, has been used
either as an adjuvant or as the decisive examination for surgical
candidate selection.11 12 13 14 15 16 One of the major
limitations of the 2-D method is that it requires mental reconstruction
of the shape of the plaque and the stenotic vessel lumen from
limited cross-sectional views. The 3-D reconstruction technique has
been investigated in carotid artery imaging employing various
modalities and has demonstrated some encouraging
findings.23 24 25 26 Whether volume-rendered 3-D
ultrasound reconstruction could allow better appreciation of the
stenotic lesions and accurate assessment of the severity of
stenosis of the cervical carotid artery is not known. Neither
is the impact of endarterectomy on the volumetric
properties of the regional carotid artery fully
understood.27 We have used a prototype 3-D
vascular ultrasound reconstruction system to evaluate its
reproducibility in the measurement of plaque length and volume and its
value in quantitative assessment of the severity of stenosis
compared with angiography. The changes in segmental vessel volume and
average cross-sectional area of the carotid artery before and after
endarterectomy were evaluated as well.
Subjects and Methods
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Patients
In this study, we recruited 14 patients (4 females; mean age, 64
years; range, 41 to 77 years), each with a recent transient
ischemic attack or minor stroke. Carotid angiography showed at
least 1 severe stenosis (70%) of the carotid artery. All
were considered candidates for carotid
endarterectomy. We performed 3-D ultrasound of both
carotid arteries within 3 months of carotid angiography or less than 3
weeks before surgery (baseline study). Thirteen of these patients
underwent carotid endarterectomy, and 3-D
ultrasound of the operated artery was repeated in 12 patients within 1
week after surgery (postoperative study). Informed consent for this
study was obtained from each patient.
Selective carotid artery angiography was performed in a standard
manner with use of the Seldinger technique. Each carotid artery was
imaged from multiple projections. All angiograms were reviewed by
an experienced independent observer to assess the site and severity of
carotid artery stenosis with the method used in the European
Carotid Surgery Trial.1 A decrease of 50% of
the estimated original lumen was defined as significant
stenosis; severe stenosis was diagnosed when the lumen
diminishment was 70%.
Carotid endarterectomy was performed in the
classic way.28 29 During the procedure, the
cerebral perfusion was monitored with electroencephalography (EEG). A
shunt was used only on indication of ischemia, when the EEG
became asymmetric. After exposure and clamping, the artery was
dissected longitudinally to expose the lumen, and the intima-media
complex of the stenotic segment was excised. After close
examination to ascertain that there was no debris left in the lumen,
the artery was primarily closed with a running suture. In case the
operated segment was of small diameter, the incision was closed with
use of a polytetrafluorethylene patch to prevent the lumen from being
too narrowed.
Instrumentation
A commercially available ultrasound system (AU3 Partner, Esaote
Biomedica, SpA) incorporated with a prototype probe was used for
3-D imaging of the carotid arteries. The prototype probe has a linear
array transducer operating at dual frequencies of 10/7.5 MHz. The
transducer is encased inside the probe, surrounded by mineral oil
(Shell Cassida Fluid HF 16) as the transmission media. The fan-like
movement of the transducer is steered by a stepper motor inside the
probe controlled by the ultrasound system. The intervals between 2D
images during data acquisition are programmable. The range of
transducer movement can be predetermined at 30° to 65°. A prism 3-D
ultrasound data set is acquired by automatic sequential collection of
images in a push-button manner, with the probe held in a fixed
position. 2-D images can be collected consecutively without ECG gating
to produce a static data set, or a dynamic data set can be obtained by
registration of ECG-gated images to their correspondent cardiac phases.
The probe also has the ability of conventional 2D and spectral and
color Doppler imaging.
In each patient, both carotid arteries were imaged before
endarterectomy, and the operated artery was imaged
after surgery. We first performed 2D ultrasound imaging and color and
spectral Doppler to select the optimal acoustic window for 3-D data
acquisition. The internal carotid artery was differentiated from the
external carotid artery by their morphology (eg, the enlargement of the
internal carotid artery at the bulb), their spatial relationship
(judged from their positions in the imaging sector with known location
of the probe), and the characteristics of the spectral Doppler from
these vessels. This information was used later as well, in 3-D data
analysis. A multitude of 2-D images of the carotid artery was
collected sequentially at 1° intervals by steering the transducer
through 65° without ECG gating. This process took less than 2
seconds. The acquired images were stored in a digital format and were
simultaneously processed by the incorporated computer
software within the ultrasound unit. The gaps between the 2-D images
were interpolated with a bilinear algorithm to produce a volumetric 3-D
data set, which was stored for off-line analysis (Figure 1
). The 3-D data set was reconstructed
with a reference system relative to the alignment of the probe. Any
cross section selected through the volumetric data set was computed
nearly in real-time. Images of the 2D ultrasound before and during 3-D
data acquisition were stored onto VHS videotapes for necessary
review.
View larger version (38K):
[in a new window]
Figure 1. Schematics showing the principle of 3-D ultrasound
data acquisition and production of a volumetric data set of the
carotid artery.
The 3-D data sets were reviewed by an independent observer
unaware of angiographic results to determine the presence or absence of
plaques and the site of stenosis. Measurement of plaque length
and volume and severity of stenosis were performed separately
by 2 blinded observers and repeated by one of them with intervals of 1
week or longer. ).
View larger version (29K):
[in a new window]
Figure 2. Schematic description of the diameter method and
the area method for measuring the severity of carotid stenosis
on the cross-sectional image reconstructed from a 3-D ultrasound data
set. ICA indicates internal carotid artery; ECA, external carotid
artery; CCA, common carotid artery; S, stenosis;
DL, minimal diameter of the free lumen; DV,
diameter of the original vessel; AL, minimal area of the
free lumen; and AV, cross-sectional area of original
vessel.
All data were expressed as mean±SD. Comparison between
measurements by 3-D ultrasound and angiography before and after surgery
and intraobserver and interobserver variability were examined using
linear regression, paired Student t test, and Bland-Altman
analysis. A value of P<0.05 was defined as
statistically significant.
Results
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Carotid Endarterectomy
Endarterectomy was successful in all 13
patients who underwent this procedure. Among the 12 patients who had
3-D ultrasound both before and after surgery, 3 required intraoperative
shunts, 3 had polytetrafluorethylene patches on closure of the
arterial incision, and 2 suffered from postoperative
bleeding that required reoperation within 24 hours after surgery. There
was no other major perioperative event.
In all 14 patients, 26 significant stenoses (5% to 100%,
82±17%) were found in 19 of 28 examined vessels by carotid artery
angiography. Eight of these stenoses were located in internal,
8 in external, and 1 in common carotid arteries, and 9 were at the
bifurcation extending from common to internal carotid artery. Of these
stenoses, 21 were severe (70%).
3-D ultrasound was performed in all 14 patients in baseline
studies and in 12 patients in postoperative studies. In baseline
studies, 3-D data sets were obtained from both left and right carotid
arteries; for the postoperative studies, 3-D data sets of the operated
carotid artery were obtained. It took less than 2 seconds to acquire
each data set.
All the significant stenoses diagnosed from angiography
were recognized from 3-D ultrasound except 1 in the external carotid
artery (due to suboptimal image quality). The severity of
stenosis measured from 3-D ultrasound ranged from 45% to 100%
(74±19%) using the diameter method and 40% to 100% (83±18%) using
the area method. The correlation between the percentage of
stenosis measured from angiogram and that measured from 3-D
ultrasound using the area method (SEE=10.5%, r=0.82, mean
difference=1.8±10.5%) was better than that using the diameter method
(SEE=12.4%, r=0.76, mean difference 7.0±12.3%)
(Figure 3). The sensitivity, specificity,
positive predictive value, and negative predictive value of 3-D
ultrasound in defining severe carotid stenosis were 65%,
100%, 100%, and 65%, respectively, using the diameter method, and
90%, 92%, 95%, and 86%, respectively, using the area method.
View larger version (37K):
[in a new window]
Figure 3. Linear regression (left) and Bland-Altman
analysis (right) between the severity of carotid
stenosis measured from angiography and that from 3-D ultrasound
using the diameter method (top) and the area method (bottom).
From 3-D ultrasound data sets of the carotid artery, the length
and volume of 20 randomly selected plaques were measured. They ranged
from 3 to 35 mm (15±8 mm) in length and 45 to 2980
mm3 (703±734 mm3) in
volume. There was excellent correlation for both intraobserver and
interobserver measurements (Figures 4A and 4B).
View larger version (36K):
[in a new window]
Figure 4. A, Comparison between intraobserver (top) and
interobserver (bottom) measurements of the lengths of 20 randomly
selected plaques. B, Comparison between intraobserver (top) and
interobserver (bottom) volume measurements of 20 randomly selected
plaques.
At the most stenotic site, the cross-sectional area of the
carotid artery free lumen changed from 6.4±6.3
mm2 before endarterectomy to
50.6±18.8 mm2 after. The segmental vessel
volume and the average cross-sectional area of the operated artery (2
cm long, including the bifurcation) changed on average from 2227
mm3 (874 to 3240 mm3)
and 111 mm2 (44 to 162
mm2) before surgery to 2318
mm3 (1648 to 3368 mm3)
and 115 mm2 (82 to 168
mm2) after surgery, with a slight but
nonsignificant increase (Figure 5). By
dividing the patients into a group with patches applied during surgery
and a group without patches, a significant increase was observed in the
segmental volume in the former group. In the latter group of patients
without patches, the segmental volume had no significant change,
although the average value slightly decreased (Figure 6).
View larger version (24K):
[in a new window]
Figure 5. Graphs showing changes in segmental vessel volume
(left) and average cross-sectional area (right) of the carotid artery
before endarterectomy (pre-EAT) and after
(post-EAT). Since the lengths of all the measured segments were the
same (2 cm), the changes in segmental volume and in cross-sectional
area are proportional.
View larger version (32K):
[in a new window]
Figure 6. Plots showing changes of segmental vessel
volume at baseline (black bars) and postoperative (gray bars) in each
individual patient (top) and in the groups of patients with (bottom
left) and without (bottom right) polytetrafluorethylene patches during
endarterectomy on closure of the artery. Patches
were applied in patients 4, 5, and 12 (underlined). The postoperative
segmental vessel volume in the group of patients with patches had
significant increase compared with baseline measurement (*).
Besides the quantitative information of the carotid artery and
stenotic lesions, we were able to obtain some incremental
information of the carotid artery, plaques, and surrounding structures
from 3-D ultrasound. Not only the longitudinal but also the
circumferential extent of the plaque was well appreciated at various
levels. The eccentricity of the plaque distribution and the
cross-sectional area and shape of the free vessel lumen were better
portrayed throughout the segment of the carotid artery that was within
the 3-D data set (Figure 7). Presence or
absence of calcification in the plaque or the vessel wall could be
predicted from the intensity of the ultrasound signal in comparison
with surrounding tissues, with the calcified plaques or vessel wall
appearing brighter and usually causing shadowing or ultrasound
attenuation.
View larger version (102K):
[in a new window]
Figure 7. Images reconstructed from 3-D ultrasound of the
carotid arteries in different patients, labeled A, B, C, and D. The
thick arrow in each panel under the capital letter indicates the
direction of carotid artery from proximal to distal. Four images are
shown in each panel as they appear on the screen of the 3-D ultrasound
unit. The upper left image shows the orientation of the volumetric data
set in a cube. A cross-sectional cutting plane of the carotid artery
can be reconstructed through guidance from a spatial coordinate system
(upper right image). At the same time, 2 other cutting planes
perpendicular to the first are displayed automatically, and their
positions are indicated by 2 vertical lines (1 and 2). Panel A shows
3-D reconstruction of a carotid artery after
endarterectomy. Panels B, C, and D are examples of
multiple plaques with different shapes and distribution (thin arrows).
The longitudinal extent of the plaques are shown in the upper right
images; the lower images display their circumferential involvement of
the vessel and the cross-sectional view of the free carotid lumen as
well as the shape of the plaques.
Discussion
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Imaging Techniques for the Carotid Artery
Among the techniques that have been used for evaluation of
carotid artery stenosis, angiography is the most widely used
and accepted for selection of candidates for carotid
endarterectomy, as it was in the North American
Symptomatic Carotid Endarterectomy
Trial (NASCET) and the European Carotid Surgery Trial
(ECST).1 2 3 4 5 6 However, carotid angiography is an
invasive procedure with nonnegligible complications and peri-procedural
risk, especially in patients with severe
stenosis.7 8 30 31 Besides, the
technique provides exclusively free lumen projections, ignoring
details of the plaques, the vessel wall, and the surrounding tissues
and structures. Therefore, the assessment of the percentage of carotid
artery narrowing at the stenotic site requires either
extrapolation of the original diameter of the vessel lumen at the same
site (ECST) or measurement of vessel diameter at a reference site,
such as the segment of the vessel distal to the stenosis
(NASCET) or the common carotid artery.1 2 3 4 5 6 32
Experience with intravascular ultrasound in carotid artery imaging is
still preliminary.33 Although it may provide
information about the arterial wall and stenotic
lesions on cross-sectional views, it gives less information on the
longitudinal extension of the lesions. Besides its invasiveness and
expense, intravascular ultrasound is limited in use in severe carotid
stenosis. MR angiography and CT angiography of the carotid
artery have been investigated as well.17 18 19 20 21 22 Both
techniques are expensive and neither is available at bedside. The
latter also requires contrast injection. Therefore, their application
is limited in daily practice. 2D ultrasound is able to overcome some of
the drawbacks of the above-mentioned techniques and has proved reliable
in some previous studies in the evaluation of carotid artery
stenosis, especially when combined with Doppler
imaging.11 12 13 14 15 16 However, its feasibility and
accuracy has been challenged by the physical inaccessibility of some
cutting planes, which in some circumstances might be crucial for
diagnosis. In addition, 2D ultrasound is technically operator dependent
in the acquisition of the useful on-line information. 3-D ultrasound
imaging of the carotid arteries has brought the attention of clinical
workers to its potential use.23 24 34
3-D surface ultrasound, realized by sequential collection of
2-D images of the carotid artery, results in a volumetric digital data
set. The method we used in this study has several advantages. First, it
is noninvasive and portable and can be performed in various clinical
settings. Second, it may minimize the discomfort of the patient by
reducing the examination time and probe manipulation. A 3-D data set
can be collected within 2 seconds, with the probe held in a fixed
position. Close examination of the carotid artery can be achieved
off-line, and images of the carotid artery can be reconstructed in
unrestricted directions from the 3-D data set. Third, it provides
volumetric information of not only the free lumen of the carotid artery
but also the plaques, the vessel wall, and the adjacent tissue and
structures, such as the jugular vein. Information of the shape and
distribution of plaques and the degree of calcification may be helpful
in clinical management of the patients, such as selection of
appropriate interventional methods. And last, 3-D ultrasound permits
volume quantification of either a plaque or a segment of free lumen or
original vessel lumen. The reproducibility of plaque length and volume
measurements in this study, both from the same observer and from
different observers, was excellent. This may have important clinical
implications in serial follow-up studies. For example, it may provide a
reliable method for serially observing the progression or regression of
a plaque and/or changes in the severity of stenosis in a
segment of the carotid artery. It may also provide a reliable method in
follow-up studies after interventional procedures, such as carotid
endarterectomy or endoluminal stenting, to observe
the local vessel change, plaque reformation, or stent dysfunction, such
as inadequate expansion or recoil.
Our study had several limitations. First, carotid
angiography was used as the reference method for quantitation of
carotid stenosis. However, it has its own potential limitations
in the accuracy of estimating a 3-D stenotic lesion by using a
2-D projection of the lumen silhouette and by assuming the original
lumen size of the vessel at the diseased site. Therefore, although in
this study the severity of carotid stenosis measured by 3-D
ultrasound using the area method and the diameter method correlated
well with the results from carotid angiography, care must be taken when
interpreting the results. Second, some problems of the 2-D ultrasound
could not be overcome by 3-D reconstruction. For instance, image
quality of 2-D ultrasound could not be improved by 3-D reconstruction.
On the contrary, interpolation of the spaces between the original 2-D
images further decreases the image resolution. Ultrasound artifacts in
2-D images will remain in the 3-D data set, which also affect the
reconstructed images. Finally, the number of patients (n=14) examined
in this study was small, although the number of vessels we studied
(n=40) was considerable. The measurement of plaque volumes could not be
validated in this study. Instead, we tested the intraobserver and
interobserver variabilities, and the results proved this method to be
highly reproducible. In addition, the plaques we observed in this study
had various sizes, shapes, and echogenicity. However, the number of
lesions was not great enough to divide them into subgroups. Therefore,
the accuracy of measurement affected by the tortuosity of the free
vessel lumen and by the morphology of the plaques was not examined.
Further investigation in a larger population with a wider range of
carotid abnormalities is necessary to validate the results from this
study.
3-D ultrasound of the carotid arteries can be used to detect
and quantify significant and severe carotid stenosis. Its
potential in volumetric measurements indicates important clinical
implications. Quantification of plaque and vessel volume allows serial
follow-up studies of the progression or regression of stenotic
lesions and evaluation of interventional procedures.
References
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
This article has been cited by other articles:
 |
|
 |
|
  W. W. Lien, A. H. Lee, Y. Kono, G. C. Steinbach, and R. F. Mattrey Noninvasive Estimation of the Pressure Gradient Across Stenoses Using Sonographic Contrast: In Vitro Validation J. Ultrasound Med., May 1, 2004; 23(5): 683 - 691. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Schminke, L. Motsch, L. Hilker, and C. Kessler Three-Dimensional Ultrasound Observation of Carotid Artery Plaque Ulceration Stroke, July 1, 2000; 31(7): 1651 - 1655. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||