IMAGINE
Maximiliano Arroyo
UT Division of Cardiovascular
Diseases
January 19th, 2006



In 1999, more than 1.83 million coronary angiograms
were performed in the US. Only 1/3rd were performed in
conjunction of an interventional procedure.
CT is the premier noninvasive modality for vascular
imaging of the thorax; the heart, however, has always
been technically challenging because of its continuous
motion.
The cross-sectional nature of CT may enable
assessment of the vessel wall. The potential for
noninvasive identification, characterization, and
quantification of atherosclerotic lesions and total disease
burden within the coronary arteries is currently being
evaluated.
U. Joseph Schoepf. Radiology 2004;232:18-37
Modalities:
 EBCT:
Introduced in 1984, was the first
system to enable ECG-synchronized CT
imaging of the cardiac anatomy. With
presently available scanners, the routine
protocol comprises a collimation of 3 mm,
a temporal resolution of 100 msec, and
prospective ECG triggering for sequential
acquisition of transverse images
consistently at the same phase of the
cardiac cycle, typically during diastole.
U. Joseph Schoepf. Radiology 2004;232:18-37
MDCT: Introduced in 1998, mechanical
spiral CT systems with simultaneous
acquisition by four detector rows and a
minimum rotation time of 500 msec were
introduced. This provided a substantial
performance increase over the spiral CT
systems that had been available until then.
U. Joseph Schoepf. Radiology 2004;232:18-37

A higher temporal resolution is enabled by
means of faster gantry rotation speed combined
with dedicated image reconstruction algorithms.
The strategy that has been pursued to further
improve fast high-resolution volume coverage is
to increase the number of sections that are
simultaneously acquired. So far, this has
resulted in the introduction of eight–, 10–, 16–,
32–, 40–, and 64–detector row CT scanners with
further reduced gantry rotation times and
minimum beam collimation widths of less than 1
mm.

Presence of severe calcification is a limitation of
contrast-enhanced CT coronary angiography because
beam-hardening and partial-volume effects can
completely obscure the cross section of the vessel and
prevent assessment of the patency of the coronary
artery lumen. Owing to similar effects, metal objects
such as stents, surgical clips, and sternal wires can also
interfere with the evaluation of underlying structures. Use
of the thinnest possible section width reduces partialvolume artifacts to some extent and improves
assessment of calcified coronary segments.
U. Joseph Schoepf. Radiology 2004;232:18-37
ECG-synchronized CT Scan Acquisition
 Prospective
Triggering:
A trigger signal is derived from the patient’s ECG
on the basis of a prospective estimation of the
present R-R interval, and the scan is started at a
defined time point after a detected R wave,
usually during diastole. With MDCT, several
sections are obtained simultaneously during one
scan acquisition with a cycle time that ordinarily
allows image acquisition at every other
heartbeat. In general, this strategy results in
shorter breath-hold times, and respiratory
artifacts are less likely to occur.
U. Joseph Schoepf. Radiology 2004;232:18-37
 To
improve temporal resolution, scan data
are only acquired during a partial scanner
rotation (approximately two-thirds of a
rotation with 240°–260° projection data),
which covers the minimum amount of data
required for image reconstruction. In this
way, prospective ECG triggering is the
most dose-efficient method for ECGsynchronized scanning.

However, only rather thick section collimation (3
mm with EBCT, 1.5 mm with 16—detector row
CT) is usually being used for a prospectively
ECG-triggered acquisition. Thus, the resulting
data sets are less suitable for 3D reconstruction
of small cardiac anatomy. Also, the prospectively
ECG-triggered technique greatly depends on a
regular heart rate of the patient and is bound to
result in misregistration in the presence of
arrhythmia.
U. Joseph Schoepf. Radiology 2004;232:18-37
 Retrospective Gating:
An alternative approach is retrospective ECG
gating. This generally enables greater flexibility
for phase-consistent image reconstruction when
examining a patient with a changing heart rate
during acquisition. Retrospective ECG gating
requires multi–detector row spiral scanning with
a slow table motion and simultaneous recording
of the ECG trace, which is used for retrospective
linkage of scan data with particular phases of the
cardiac cycle.
U. Joseph Schoepf. Radiology 2004;232:18-37
 Retrospectively
ECG-gated CT of the
heart requires a highly overlapping spiral
scan with a spiral table speed adapted to
the heart rate to ensure complete phaseconsistent coverage of the heart with
overlapping image sections.
At heart rates less than a predefined threshold, one segment of consecutive
multisection spiral data is used for image reconstruction. At higher heart rates,
two or more subsegments from adjacent heart cycles contribute to the partial
scan data segment. In each cardiac cycle, a stack of images is reconstructed at
different z-axis positions covering a small subvolume of the heart .
U. Joseph Schoepf. Radiology 2004;232:18-37



The continuous spiral acquisition enables reconstruction
of overlapping image sections, with a longitudinal spatial
resolution of up to 0.6 mm.
Retrospectively ECG-gated acquisition is the preferred
method for contrast-enhanced high-spatial-resolution
imaging of small cardiac structures, especially the
coronary arteries.
Diastole is usually chosen for image reconstruction
because it is the phase of the cardiac cycle with the least
motion; however, owing to the highly overlapping
acquisition, image data can be reconstructed for the
entire course of the cardiac cycle.
U. Joseph Schoepf. Radiology 2004;232:18-37
 Optimizing

Spatial Resolution
Spatial resolution is largely dependent on the type of
scanner available. The smallest detector widths range
from 0.5 to 1.25 mm.
The spatial resolution of four detector row CT is 0.6 x 0.6
x 1.0 mm, that of electron-beam CT is 0.7 x 0.7 x 3 mm,
and that of magnetic resonance (MR) coronary
angiography is 1.25 x 1.25 x 1.5 mm. Spiral CT allows
volume acquisition and reconstruction of overlapping
sections, which improve z-axis resolution. The resolution
of 16 detector row CT is up to 0.5 x 0.5 x 0.6 mm. This
resolution is approaching, but remains inferior to, that of
conventional angiography, which is 0.2 x 0.2 mm.
Harpreet P.Radiographics. 2003;23:S111-S125
 Optimizing
Temporal Resolution
The temporal resolution is the amount of time it
takes to acquire the necessary scan data to
reconstruct an image. The temporal resolution of
electron-beam CT is 100 msec, and that of MR
imaging is 100–150 msec. For multisection CT, it
is primarily dependent on the time taken by the
scanner to complete one gantry rotation but can
be modified by using partial scan reconstruction
techniques.
Harpreet P.Radiographics. 2003;23:S111-S125
Radiation Dose

Relatively high radiation exposure is involved
with retrospectively ECG-gated imaging because
of continuous x-ray exposure and overlapping
data acquisition at a slow spiral table feed, a
substantial portion of the acquired data and
radiation exposure are redundant and do not
contribute to image generation.
 There is considerable disagreement in the
literature as to the actual radiation dose,
because the lack of standardization of the
protocols.
U. Joseph Schoepf. Radiology 2004;232:18-37

For high spatial resolution (1.00–1.25-mm beam
collimation), a retrospectively ECG-gated
acquisition), and routine scanner settings with
four–detector row CT, an exposure limit of
approximately 10 mSv is applied, which is two to
three times the average annual background
radiation in the United States. Comparable to the
exposure received during a typical routine
diagnostic coronary angiogram. As progressively
thinner beam collimations are used for scanner
types with added detector rows, radiation dose
generally increases.
CR Conti. Clin. Cardiol. 28:450-453
CR Conti. Clin. Cardiol. 28:450-453
Contrast Injection


Scanning times for imaging of the heart with 8 or 16
detector row CT scanners range from 20 to 40 seconds,
80–120 mL contrast medium injected at a rate of 3–5
mL/sec is needed to maintain homogeneous vascular
contrast throughout the scan.
Saline chasing (eg, bolus of 50 mL of saline injected
immediately after the iodinated contrast medium bolus)
has proved to be helpful for better contrast medium
bolus utilization, for high and consistent vascular
enhancement, and for prevention of streak artifacts,
which frequently arise from dense contrast material in
the superior vena cava and right atrium and sometimes
interfere with the evaluation especially of the right
coronary artery.
U. Joseph Schoepf. Radiology 2004;232:18-37
Data Display
 Maximum intensity projection: Not only
displays coronary artery CT data in a more
intuitive format but also condense diagnostic
information into a few relevant sections or views.
For routine visualization of large-volume CT
coronary angiography data sets, many centers
perform three dedicated maximum intensity
projection reconstructions to create views of the
left and right coronary arteries and of the entire
coronary arterial tree from a cranio-oblique
perspective.
U. Joseph Schoepf. Radiology 2004;232:18-37



(a) RAO view along the interventricular groove
shows LAD, with mixed atherosclerotic lesion
(arrowhead) with calcified components in the
proximal course of the vessel.
(b) LAO view in plane RCA with calcified nodules
(arrowheads) along the course of the vessel.
(c) LAO "spider" view shows (LAD and its diagonal
branches, with soft-tissue-attenuation plaque
(arrowhead) in the anterior aspect of the left main
coronary artery (LM) wall.
U. Joseph Schoepf. Radiology 2004;232:18-37
 Multiplanar
reformations: image
data can be
rearranged in
arbitrary imaging
planes, with image
quality comparable to
that of the original
transverse sections.
left anterior descending coronary artery in a
patient with CAD.
U. Joseph Schoepf. Radiology 2004;232:18-37
 Three-
dimensional
display: 3D post
Left: Anteroposterior cranial projection shows
LAD and Cx.
Right: Volume rendering in anteroposterior
cranial projection shows left main coronary artery
with its branches, LAD and Cx.
processing is a means of
displaying information in
an intuitive fashion. The
most commonly used
technology for 3D display
of the coronary arterial
tree is volume rendering.
U. Joseph Schoepf. Radiology 2004;232:18-37

Contrast-enhanced 16-detector
row CT coronary angiography.
Colored volume rendering of
right coronary artery (RCA)
displayed in slightly cranial
right anterior oblique.
U. Joseph Schoepf. Radiology 2004;232:18-37
Contrast-enhanced CT of Coronary Artery
Anomalies, Bypass Grafts, and Stents
 MR
imaging is limited with regard to
determination of the distal coronary arterial
course. Therefore, CT is the preferred
modality for evaluation of small collateral
vessels, fistulas, and vessels originating
outside the normal sinuses.
U. Joseph Schoepf. Radiology 2004;232:18-37
Patient with superdominant anomalous right coronary artery (AnRCA) supplying the majority of the
myocardium. (a) Selective conventional angiographic image and (b) volume-rendered 3D
reconstruction (cranial right anterior oblique perspective) from contrast-enhanced 16-detector row CT
coronary angiography.
U. Joseph Schoepf. Radiology 2004;232:18-37
 Bypass
graft imaging: more clinically relevant, is
complex functional assessment of bypass flow, accurate
detection of graft lesions, and reliable visualization of
(distal) anastomoses. Data on the accuracy of CT for the
detection and grading of hemodynamically significant
graft stenosis are still rather sparse and are ordinarily
based on small patient populations studied with electronbeam or multi–detector row CT.
In a somewhat larger patient population investigated with
four–detector row CT, overall sensitivity and specificity
values for bypass occlusion of 97% and 98%,
respectively, were reported.
1

2
1.U. Joseph Schoepf. Radiology 2004;232:18-37
2. Ropers D. Am J Cardiol 2001; 88:792-795
Thomas Schlosser. JACC 2004; 44:1224-1229

All IMA grafts could be visualized with diagnostic
image quality, whereas only 28 of 37 (76%) of
the distal anastomoses to the LAD and 3 of 5
(60%) of the distal anastomoses to the diagonal
branches could be evaluated.
 A total of 11 of 42 (26%) of the distal IMA
anastomoses were classified as unevaluable
due to poor opacification and artifacts caused
by metal clips.
Thomas Schlosser. JACC 2004; 44:1224-1229




MSCT permitted visualization of all proximal and distal anastomoses
of venous grafts to the LAD.
Invasive coronary angiography revealed 8 venous grafts to the LCX
to be occluded, all correctly diagnosed by MSCT. All proximal and
25 of 33 (76%) distal anastomoses in the LCX region were
adequately seen on MSCT. The remaining 8 distal anastomoses
(24%) were classified as unevaluable due to poor opacification
and/or artifacts caused by cardiac motion.
All proximal and 22 distal anastomoses (63%) to the RCA, could be
visualized. A total of 13 of 35 (37%) of the distal anastomoses were
classified as unevaluable. Overall, 83 of 112 (74%) distal
anastomoses could be evaluated.
The unevaluable distal anastomoses were estimated as stenotic.
This results in a lower specificity (68%) and positive predictive value
(PPV) (37%) compared with the separate analysis of the evaluable
segments (specificity 95%, PPV 81%).
Thomas Schlosser. JACC 2004; 44:1224-1229
Colored volume-rendered view from anterior
perspective, derived from 16-detector row
CT angiography, 3 venous bypass grafts
VCABG-LAD, VCABG-Cx, and VCABGRCA. Additional left internal mammary
artery bypass graft (LIMA-BG), also to the
LAD
(LIMA) bypass graft. Anastomosis has been created
between left internal mammary artery and left anterior
descending coronary artery (LAD) territory. Note
extensive atherosclerotic changes in the native
vessels.
U. Joseph Schoepf. Radiology 2004;232:18-37
 Coronary
stents have been notoriously difficult
to assess with CT. Contrast-enhanced CT can
be used to assess stent patency on the basis of
contrast enhancement in the course of the artery
with the stent, because an unenhanced distal
coronary artery lumen usually reflects critical instent restenosis. However, assessment of the
stent lumen for nonocclusive in-stent restenosis
due to neointimal hyperplasia remains
challenging.
U. Joseph Schoepf. Radiology 2004;232:18-37
(a) Colored 3D volumerendered view from right
posterior oblique
perspective reveals luminal
narrowing (arrowhead) of
artery proximal to the stent.
(b) Maximum intensity
projection and (c)
multiplanar reformation in
oblique coronal planes
show patent stent lumen
and mixed atherosclerotic
lesion (arrow) with calcified
and noncalcified
components as the cause
of stenosis proximal to the
stent. (d) Conventional
angiographic image in left
anterior oblique projection
confirms stent patency and
presence of stenosis.
U. Joseph Schoepf. Radiology 2004;232:18-37
Box and whisker plot (median value and quartiles) of angiographic in-segment
coronary stenosis (measured by quantitative coronary angiography [QCA]) for
each of the four grades of MDCT narrowing. Grade 1, none or minimal
narrowing; grade 2, moderate but obstructing <50% of the lumen; grade 3,
significant (≥50%) but not severe narrowing; grade 4, severe narrowing to total
occlusion of stented segment.
Tamar Gaspar, JACC 2005; 46: 1573-1579
 Five
stents that were not assessable by
MDCT were excluded.
 “… MDCT excluded restenosis in twothirds of patients. … this would result in
only 1 in 10 stents with restenosis being
missed (or 13.5% of patients).”
Tamar Gaspar, JACC 2005; 46: 1573-1579
Contrast-enhanced CT Angiography for
CAD
 In
763 coronary segments, CCA detected
a total of 75 lesions ≥50%.
 The MSCT correctly assessed 54 of these.
Twenty-one lesions were missed or
incorrectly underestimated. Sensitivity was
72%, specificity 97%.
Axel Kuettner. JACC 2004; 44:1230-1237
Ricardo C. Cury. AJC. 2005; 96:784-787

64 slice MSCT compared to QCA for
quantification of lesion severity
 935 of 1,065 segments (88%) could be
analyzed either quantitatively or qualitatively.
Of these, 773 of 935 (83%) segments could
be quantitatively measured by both MSCT
and QCA. Of these, 130 of 773 (17%) had
stenoses.
 Comparing the maximal percent diameter
luminal stenosis by MSCT versus QCA.
Bland-Altman analysis demonstrated a mean
difference in percent stenosis of 1.3 ± 14.2% .
Gilbert L. Raff . JACC 2005; 46:552-557

(A) Volume rendering
technique demonstrates
stenosis of right coronary
artery below the acute
marginal branch as well as
nodular coronary calcifications
largely extrinsic to the right
coronary lumen and (B) normal
left coronary artery. (C, D)
Maximum-intensity projection
of the same arteries
demonstrates severe soft
plaque stenosis of the right
coronary artery and superficial
calcific plaque. (E, F) Invasive
coronary angiography of the
same arteries
Gilbert L. Raff . JACC 2005; 46:552-557
Overall, 935 of 1,065 (88%) segments could be interpreted, 773 of 935 (83%)
quantitatively and 162 of 935 (17%) qualitatively only.
Gilbert L. Raff . JACC 2005; 46:552-557





IVUS in 38 vessels in 20 patients.
A total of 365 sections were available for the comparison with IVUS; in 161
of these (26 vessels), atherosclerotic plaques were present.
64-slice CT enabled a correct detection of plaque in 54 of 65 (83%) sections
containing noncalcified plaques, 50 of 53 (94%) sections containing mixed
plaques, and 41 of 43 (95%) sections containing calcified plaques, resulting
in an accuracy of 90% to detect any plaque (145 of 161).
In 192 of 204 (94%) sections, atherosclerotic lesions were correctly
excluded. In addition to the ability to classify calcified, mixed, and
noncalcified lesions, 64-slice CT enabled the visualization of lipid pools in 7
of 10 (70%) sections and enabled us to identify a spotty calcification pattern
in 27 of 30 (90%) sections.
In three sections without evidence for echolucency on IVUS, hypodense
areas (lipid cores) were identified by 64-slice CT. In 314 of 365 sections
(86%), consensus between IVUS and 64-slice CT was achieved regarding
the morphologic classification. The plaque type was misclassified by 64slice CT in 23 of 145 atherosclerotic sections.
Alexander W. Leber . JACC 2006 IN PRESS
Why not MRI?

16-MDCT offers better visualization of the
coronary arteries than MR.
 Using visual assessments of DS severity, both
MDCT and MR have similar accuracy for
detecting significant coronary artery disease.
 Quantitative assessment of DS severity
significantly improves the diagnostic accuracy of
MDCT, but not that of MR, as compared to visual
analysis alone.
 Using quantitative assessment of DS severity,
MDCT has significantly higher diagnostic
accuracy than MR.
Joëlle Kefer. JACC 2005; 46:92-100
 By
visual analysis, MR and MDCT had
similar sensitivity (75% vs. 82%, p = NS),
specificity (77% vs. 79%, p = NS), and
diagnostic accuracy (77%, vs. 80%, p =
NS) for detection of >50 % DS.
Joëlle Kefer. JACC 2005; 46:92-100




Typical examples of
reformatted magnetic
resonance (MR) (left panels),
and multidetector row
computed tomography (MDCT)
(center panels) and
corresponding quantitative
coronary angiography (QCA)
images (right panels)
(A) Normal right and left
coronary arteries by MR,
MDCT, and QCA.
(B) Isolated mid-RCA stenosis.
(C) Two-vessel disease
involving the mid-LAD, and left
circumflex coronary artery
Joëlle Kefer. JACC 2005; 46:92-100
Conclusions

MDCT is now more comparable to QCA, with
excellent sensitivity and specificity in
experienced centers.
 Further evolution of MDCT (more and faster
detectors, software improvement) will likely
provide a better spatial and temporal resolution.
 Currently, MDCT is not the test of choice in
patients with prior CABG, stents, severely
calcified lesions; perhaps also patients with
elevated HR, and obese.


MDCT does not have the capability of assessing the
distribution of various morphologic patterns of calcium
and their relation to other “soft” plaque components;
further plaque characterization (e.g., lipid pools and
fibrous tissue), a prerequisite for the identification of
most vulnerable lesions, is not yet a workable reality,
even with the 64-slice machines in their current
configuration.
The noninvasive identification of plaque components
subtending vulnerable lesions will require additional
improvement in the primary instrumentation, software,
perhaps ?? the use of hybrid constructs (e.g., with
positron emission tomography).
The “ sensation 64”
Thanks!