Circulation. 2000;102:e6-e10
(Circulation. 2000;102:e6.)
© 2000 American Heart Association, Inc.
Circulation Electronic Pages |
Coronary Artery Fly-Through Using Electron Beam Computed Tomography
Peter M. A. van Ooijen, MSc;
Matthijs Oudkerk, MD, PhD;
Robert J. M. van Geuns, MD;
Benno J. Rensing, MD, PhD;
Pim J. de Feyter, MD, PhD
From the Department of Radiology (P.M.A.v.O., M.O.) and Thoraxcentre
(R.J.M.v.g., B.J.R., P.J.d.F.), University Hospital Rotterdam/Daniel,
Rotterdam, The Netherlands.
Correspondence to Peter M.A. van Ooijen, MSc, Dept of Radiology, University Hospital Rotterdam/Daniel, Groene Hilledijk 301, 3075 EA Rotterdam, The Netherlands. E-mail ooijen{at}radh.azr.nl
 |
Abstract
|
|---|
Background—Virtual reality
techniques have recently been
introduced into clinical medicine. This
study examines the possibility
of coronary artery fly-through
using a dataset obtained by noninvasive
coronary angiography
with contrast-enhanced electron-beam computed
tomography.
Methods and Results—Ten patients were examined, and 40 to
60 transaxial tomograms (thickness, 1.5 mm; in-plane pixel
dimensions, 
0.5x0.5 mm) were obtained after
intravenous contrast injection. The datasets were processed
on a graphics workstation using volume-rendering software. For
fly-throughs, the contrast-enhanced lumen was made transparent and
other tissue was made opaque. Then, key frames were selected in a path
through the vessel, with software interpolation of frames between key
frames. A typical movie contained 150 to 300 frames (10 to 15 key
frames). Fly-throughs of coronary bypass grafts (n=3), left
anterior descending arteries (LAD; n=6), and the intermediate branch
(n=1) were reconstructed. Coronary calcifications were seen in
3 patients. The fly-through of the intermediate branch, the bypass
grafts, and one of the LADs did not show any irregularities. In 2
cases, a stenosis was visible in the LAD; its presence was
confirmed by conventional coronary angiography.
Conclusions—Recent developments in fast-volume rendering
using special-purpose hardware in combination with noninvasive
coronary angiography with electron beam computed tomography
have provided the possibility of performing coronary artery
fly-throughs.
Key Words: angiography • tomography, x-ray computed • computers • imaging
|
Introduction
|
|---|
Coronary angiography is the standard way of
visualizing the
coronary arteries. However, this method is
invasive and, in
a few cases, the procedure is associated with
complications
and has a small risk of mortality. Therefore, noninvasive
methods
to visualize the coronary arteries are currently under
investigation;
these include MRI
1 2 and electron-beam
computed tomography
(EBT).
3 4 Coronary angiography
depicts the coronary artery
as a planar silhouette (lumenogram)
and, therefore, only detects
a stenosis if the plaque obstructs
the lumen; it also provides
no information on the vessel wall.
Coronary artery fly-through
is another way to provide a
comprehensive delineation of the
lumen and the impact of vessel wall
disease on the lumen. Some
examples of fly-throughs of the
coronary arteries with calcifications
and stenoses are
shown in this article.
|
Methods
|
|---|
For this study, 10 patients were selected who had a
noninvasive,
contrast-enhanced EBT angiogram of adequate quality. The
data
acquisition was performed on an Evolution XP (Imatron) EBT
scanner.
The acquisition of the 3D dataset began with the injection of
120
to 180 mL of contrast medium at 3 to 4 mL/s through an antecubital
vein.
Scanning commenced just proximal to the left main
coronary artery
after an ECG trigger at 80% of the RR interval
(diastasis).
The tomogram (slice) thickness was set at 1.5 mm, and
the table
increment after each tomogram was set at 1.5 mm,
which resulted
in contiguous, nonoverlapping slices. A total of 40 to
60 transaxial
tomograms were made during a single breath-hold.
Breath-holding
is necessary during data acquisition to avoid
respiratory motion
artifacts. Field-of-view size was generally set at
18 cm, with
a matrix size of 512
x512 pixels; this yielded a pixel size
of
0.35
x0.35 mm. The acquired data were then transferred to a
special-purpose
graphic workstation (Indigo2, Silicon Graphics, Inc)
running
VoxelView software (Vital Images) for volume rendering and
processing.
5 6 7 8
The volume dataset consisted of voxels (3D pixels), each of which
had a certain value that was based on the tissue density value measured
by the EBT scanner. Using these voxel values, several types of
renderings can be performed. One of the possibilities is to construct a
fly-through movie. A fly-through is similar in some ways to a
flight-simulator. The surroundings in a flight-simulator are all
virtually stored in a large computer database, and the image shown to
the pilot is based on the position and direction of the virtual
airplane in the virtual surroundings. By a fast and smooth replacement
of the image of the virtual surroundings, the illusion of flying is
created. In the case of a coronary artery fly-through, the
surroundings are the scanned data, and the airplane can be thought of
as a camera mounted on the tip of a catheter. The images shown are
based on the position and direction of this catheter in the
coronary artery. By displaying images at consecutive positions
along a certain path through the coronary (the flight path),
the illusion of moving through this artery is created.
To make a fly-through movie like this, the vessels must be
"hollowed out" by assigning voxels representing
contrast-media–rich blood an opacity of zero (full transparency)
(Figure 1
). Next, the viewpoint is moved
inside the aorta or coronary artery; this will be the first key
frame. After this, a number of viewpoints can be selected, which are
positioned along the flight path as key frames for the movie (typically
10 to 15 key frames are selected). These key frames and the desired
number of frames to be interpolated between the key frames are fed into
the VoxelAnimator software, which is used to render the movie. From
this information, the software interpolates a curve through the defined
key frames and renders the requested number of new frames between the
key frames along this curve, which results in a 150- to 300-frame
movie.
View larger version (87K):
[in this window]
[in a new window]
|
Figure 1. Top, these graphs contain 2 types of information.
The histogram shows the distribution of the voxel values. The
x axis represents the voxel value, and the
y axis, the number of voxels with this value. The line
graph shows the opacity value assigned to a certain voxel value. The
x axis again represents the voxel value, and now
the y axis shows the opacity percentage (low value is
transparent; high value is opaque). The images on the bottom
demonstrate the effect of a change in opacity setting. The bottom left
image shows the typical opacity setting used to depict the
contrast-rich blood in the lumen of the vessel (maximum opacification),
which is not necessarily an optimal setting for the depiction of the
coronary arteries. In this case, the soft tissue and fat (with
a lower voxel value) are transparent, and the contrast-enhanced blood
(with a higher voxel value) is fully opaque. On the bottom right, the
opacity for the soft tissue and fat is high (small peak in the yellow
curve), and the voxel values corresponding to contrast-enhanced blood
are fully transparent. As can be appreciated from the image shown, this
type of setting hollows out the vessels, which enables flying
through.
|
|
|
Results
|
|---|
The 3D EBT datasets of 39 patients were evaluated, with
special
attention paid to interslice correlation quality, lack of
artifacts,
and slice image quality. The best 10 datasets were selected
for
coronary artery fly-throughs. From these 10 patients,
fly-through
movies of coronary artery bypass grafts (n=3), the
left anterior
descending artery (LAD; n=6), and the intermediate branch
(n=1)
were constructed. Typical examples from a bypass (Figure 2
and
Figure
I [Figures I to III can be
found online at
http://circ.ahajournals.org/cgi/content/full/102/1/DC1])
and from the
LAD (Figures 3
and 4
and Figures II and III) are
shown.
Coronary calcifications were visible in 3 patients (eg,
Figure 4
and Figure
III), and a significant stenosis, which
was
confirmed by conventional diagnostic coronary
angiography,
was depicted in 2 patients (eg, Figure 3
and Figure
II).
View larger version (76K):
[in this window]
[in a new window]
|
Figure 2. A, Volume renderings of a patient with a
coronary artery bypass graft. The left image shows the artifact
introduced by the segmentation of the sternum and sternal wires as a
bright white portion of the bypass graft. Evaluating the patency of the
graft at this particular point is difficult. B, When flying through the
vessel, it is clear from the first 2 images that the graft is patent.
However, a possible calcified region shows up in the second image. By
setting the opacity curves slightly differently (third and fourth
images), the walls disappear and it becomes clear that this is not a
calcified plaque, but an artifact from the sternum and the sternal
wires. The actual fly-through is shown in Figure I.
|
|
View larger version (107K):
[in this window]
[in a new window]
|
Figure 3. A, The left image shows a 3D volume
rendering of the coronary arteries. The stenosis can be
seen in the proximal part of the LAD right after the left main. This
stenosis was confirmed using conventional coronary
angiography shown on the right. B, Some interesting frames from
the fly-through. The trifurcation through the lumen is
approached in frames 1 through 6, and movement is toward the stenosed
origin of the LAD. Frame 7 shows the stenosis in close-up and,
in frame 8, the stenosis has been crossed and the remaining
part of the LAD is seen. The actual fly-through is shown in Figure
II.
|
|
View larger version (63K):
[in this window]
[in a new window]
|
Figure 4. When flying through a calcified LAD, the
calcifications show up as white blobs floating inside the vessel. In
this case, 3 big calcifications were passed on the fly-through of the
LAD. The actual fly-through is shown in Figure III.
|
|
|
Discussion
|
|---|
EBT coronary angiography has emerged as a
potentially viable
technique for noninvasive visualization of
coronary arteries
and coronary bypass
grafts.
3 4 However, although the technique
is reasonably
robust, only 81% of the major coronary artery
branches could
be visualized with sufficient quality to assess
patency, the presence
of a severe stenosis, or total occlusion.
4 For the
reconstruction of fly-through movies, the image quality
of the 3D
datasets must be perfect. Even small irregularities
may hamper the
successful construction of a coronary artery
fly-through.
Aquisition problems include the following. (1)
Problems with
breath-holding may reduce the continuation of
a coronary artery
from one slice to another. (2) Arrhythmia,
or even a single
premature complex, may lead to images that
are triggered at a slightly
different time in the heart cycle,
resulting in a displacement of 1 to
2 mm of a single slice with
respect to the other slices. This
creates a discontinuation
of the coronary arteries in 3D
reconstructions. (3) Vessels
with a diameter <1.75 mm (area, 5
x5
pixels) will not provide
a smooth coronary fly-through.
(4) Movement artifacts of the
right coronary artery during
the 100-ms image acquisition time
hamper the construction of a
fly-through movie of this artery.
In addition to these acquisition problems, the reconstruction of
a coronary artery fly-through is very time-consuming;
therefore, only the best quality datasets with a large vessel diameter
were selected to undergo this procedure. For these reasons, a
high-quality coronary artery fly-through was reconstructed in
only 25% of the eligible patients. Much of the difficulties stated
here can be overcome in the future by improvements in the spatial and
temporal resolution of the EBT scanner. An update of the scanner is
already available with a higher spatial resolution; this will increase
the image quality and provide the possibility to display and
fly-through smaller vessels. A reduction of slice acquisition time to
below 100 ms will decrease the artifacts introduced by the movement of
the right coronary artery. New scanning techniques that allow
the acquisition of >1 slice during each heartbeat will shorten total
scanning time and, thus, reduce the artifacts introduced by
arrhythmias and breath-holding problems. Furthermore, the rapid
development of both special-purpose rendering hardware and software
will provide faster and more interactive ways to reconstruct
fly-through movies.
Fly-through movies of venous bypass grafts are relatively easy to
make because the vessel diameter is relatively large, and the cardiac
motion of these vessels is limited. However, surgical clips or sternal
wires may sometimes degrade the images because of the bright artifacts
they cause (Figure 2 and Figure I). Calcifications of the vessel wall,
which have a very high voxel value, are retained in the fly-throughs
and are visible as white blobs floating in the artery (Figure 4
and Figure III).
These preliminary findings demonstrate the feasibility and
potential of this method in coronary artery or bypass graft
fly-through movies. The technique of coronary artery
fly-through cannot be considered an alternative to traditional
coronary angioscopy because it does not provide any information
about the color of the lumen or its contents, such as plaque and
thrombus. Coronary artery fly-through is an alternative way to
evaluate noninvasive coronary angiography, and it has several
advantages. (1) It provides a delineation of the "true" 3
dimensions of the vessel lumen, unlike diagnostic
angiography (lumenography), which is limited by foreshortening and
overlapping structures. (2) Fly-throughs may eliminate the
time-consuming segmentation of overlapping, obscuring anatomical
structures (left atrium, coronary sinus) that is needed to
visualize the coronary arteries from the outside. (3)
Fly-throughs may provide a more comprehensive delineation of
bifurcation lesions or anastomoses of grafts on native vessels, which
are sometimes difficult to asses, even with routine
diagnostic angiography. Finally, fly-throughs may be
helpful in assessing the remaining coronary lumen of a heavily
calcified coronary plaque or stented segment, which may be
invisible with traditional EBT-derived 3D rendering techniques.
Conclusions
A fly-through of coronary arteries and venous bypass
grafts is feasible in clinical practice and may represent a
future diagnostic technique that will allow comprehensive
3D delineation of the vessel lumen.
|
Footnotes
|
|---|
Figures I through III can be found at http://circ.ahajournals.org/cgi/content/full/102/1/DC1
Received December 1, 1999;
revision received January 25, 2000;
accepted February 7, 2000.
|
References
|
|---|
-
Manning WJ, Li W, Edelman RR. A preliminary report
comparing magnetic resonance coronary angiography with
conventional angiography. N Engl J Med. 1993;328:828–832.[Abstract/Free Full Text]
-
Van Geuns RJM, de Bruin HG, Rensing BJWM, et
al. MRI of the coronary arteries: clinical results
from three-dimensional evaluation of a respiratory gated technique.
Heart.. 1999;82:515–519.[Abstract/Free Full Text]
-
Moshage W, Achenbach S, Seese B, et al.
Coronary artery stenoses: three-dimensional imaging
with electrocardiographically triggered, contrast agent-enhanced,
electron-beam CT. Radiology. 1995;196:707–714.[Abstract/Free Full Text]
-
Rensing BJ, van Geuns RJM, Bongaerts AHH, et al.
Intravenous coronary angiography using electron
beam computed tomography, a viable alternative to conventional
coronary angiography. Circulation.. 1998;98:2509–2512.[Abstract/Free Full Text]
-
Ney DR, Fishman EK, Magid D, et al. Volumetric
rendering of computed tomography data: principles and techniques.
IEEE Computer Graphics Appl. 1990;10:24–32.
-
Fishman EK, Magid D, Ney DR, et al. Three-dimensional
imaging: state of the art. Radiology. 1991;181:321–337.[Free Full Text]
-
Heath DG, Soyer PA, Kuszyk BS, et al. Three
dimensional spiral CT during arterial portography:
comparison of three rendering techniques.
Radiographics. 1995;15:1001–1011.[Abstract]
-
Kuszyk BS, Heath DG, Ney DR, et al. CT
angiography with volume rendering: imaging findings. AJR Am
J Roentgenol. 1995;165:445–448.[Free Full Text]
Top
Abstract
Introduction