Cardiac Robert
R.
Edelman,
MD
Coronary
#{149} Warren
acquired
per
segment.
By using
overlapping 4-mm-thick sections, the coronary arteries were routinely depicted from the coronary ostia distally at MR in healthy subjects. Ultrafast MR angiography of the coronary arteries is feasible with use of a standard body coil. This technique offers considerable potential as an investigational tool and, with further development, may become ful imaging application. Index 54.1214 studies,
Manning,
Arteries:
The authors describe a method for performance of ultrafast magnetic resonance (MR) angiography of coronary arteries with a standard clinical MR system and a body coil. Each image was obtained within a single breath hold by using an electrocardiography-gated, segmented, ultrafast, gradient-echo pulse sequence with an incremental excitation flip angle for the eight phase-encoding steps
J.
a clinically
use-
terms: Coronary vessels, MR studies, #{149} Magnetic resonance (MR), vascular 54.1214 1991;
181:641-643
M
raphy
in several
clinical
ing screening rysms (1) and the extracranial
PhD
(MR) being
applications,
system,
teries (3). Because
of the
giography
in these
includ-
and
success
renal
of MR
other
of
aran-
applications,
techniques.
artery spatial
Moreover,
coro-
flow is phasic (4), and resolution is needed to
The high achievement
MD,
PhD
in rapid
sequence,
constituting
one segment (6). One segment quired in diastole during each
was acheartbeat
by using
triggering,
electrocardiographic
which breath
permitted hold.
imaging
within
Initially, findings with several variations of the pulse-sequence parameters were compared. These variations included the absence or use of flow-compensated gradients with the concomitant alteration in echo times, the use of a fixed versus an incremental excitation flip angle (7), and
the use of prepulses background
signal
comparisons
were
for suppression (8). After made,
of
the initial
most
imaging
was performed with a flow-compensated, fast-imaging-with-steady-state-precession sequence, with repetition time of 14 msec, echo time of 8 msec, section thickness of 4
separate the signal intensities of arteries from those of surrounding fat, myocardium, and blood in cardiac
chambers. requires
acquired
seconds, a single
vivo coronary artery imaging. Because respiratory motion causes blurring, cardiac gating is necessary; therefore, image acquisition within a breath hold is impossible with use of nary high
Paulin,
with 16 segments interleaved in order to complete a 128 x 256 matrix in 16 heartbeats. Imaging time was typically 12-16
we sought to image the coronary arteries, which typically are similar in caliber to intracranial vessels. There are several obvious challenges for in
standard
#{149} Sven
The imaging technique consisted of a two-dimensional, flow-compensated, gradient-echo sequence modified in such a way that eight phase-encoding steps were
angiogused
for intracranial aneudetection of stenoses carotid arteries (2),
vertebrobasilar
mm,
spatial resolution of an adequate
field
of 128
of view
x 256
(16
of 270
x 270
segments
mm,
of eight
matrix phase-
gradient-echo acquisitions (5). We have modified this approach in human studies to permit acquisition of electrocardiography-gated coronary artery images in a body coil within a
encoding steps), with an incremental flip angle series of 10#{176}, o 350 50#{176}, 50#{176}, 50#{176}, 50#{176}, 50#{176}. A scout view was first obtained in the coronal plane, after which transverse images were obtained at the level of the coronary ostia. Sequential transverse images were obtained with 1-mm overlap. In some cases, oblique views were obtained from the transverse images once the proximal coronary artery had been identified. Twenty-two healthy adult volunteers were studied (16 men and six women,
single
aged
of signal-to-noise
difficult
with
onary
ratio,
a body
which
coil.
of in vivo
with
breath
rodent
segmented,
hold.
The
hearts
ultrafast,
method
was
applied in a series of healthy subjects to determine the feasibility of imaging human coronary arteries. SUBJECTS
AND
as a radio-frequency
23-68
years).
formed within Israel Hospital
transmitter
gradient amplitude ramp time to peak
and rewas 10 amplitude
All
studies
were
per-
the guidelines of the Beth Committee on Clinical In-
vestigations.
METHODS
Images were obtained with a 1.5-T whole-body imaging system (Magnetom SP; Siemens Medical Systems, Iselin, NJ) with a circularly polarized body coil used ceiver. Peak mT/rn, with of I msec.
is
published articles have that imaging of the cor-
arteries
is feasible
the Departments of Radiology (R.R.E., D.B., S.P.) and Medicine (Cardiovascular Division) (W.J.M.), Charles A. Dana Research Institute and the Harvard Thorndike Laboratory of the Beth Israel Hospital and Harvard Medical School, Boston. Received May 3, 1991; revision requested June 10; revision received July 8; accepted July 16. Address reprint requests to R.R.E., Department of Radiology, Beth Israel Hospital, 330 Brookline Aye, Boston, MA 02215. C RSNA, 1991 See also the editorial by Caputo (pp 629-630) in this issue.
Burstein,
MR Anglography’
resonance is currently
AGNETIC
Previously demonstrated
1From
#{149} Deborah
Breath-hold
level Radiology
MD
Radiology
RESULTS Initial
studies
consisted
of a com-
parison of findings with use of various pulse-sequence parameters. In all cases, a fast-imaging-with-steadystate-precession sequence was used
rather quence
than
a fast-low-angle-shot
to avoid
the
occurrence
seof
artifacts
at high
excitation
flip
angles
(9). Use of an 8-msec echo flow compensation yielded
time with better im-
age
echo
quality
than
of 6 msec In general,
use
of an
time
without flow compensation. use of the incremental
ex-
citation flip angle series gave better signal-to-noise ratio of coronary artery flow than did use of a fixed flip angle of 15#{176} or 25#{176} and thereby per-
mitted Use
use
of a smaller
of larger
fixed
field
flip
of view.
angles
up
to
tended to suppress the signal from in-plane flow. Use of a prepulse to 500
suppress necessary
stationary to obtain
of coronary not
artery
generally
spins was not sufficient contrast
flow,
and
it was
By using
the
segmented
fast-imag-
ing-with-steady-state-precession quence and the incremental
seexcitation
flip angle series, portions mal left and right coronary were
seen
routinely
The resting to 84 beats imaging left and
well
varied from 46 resulting in an
of 11-21 coronary
delineated
of the proxiarteries
seconds. arteries
by using
The were
a transverse
plane of section (Figs 1, 2). Oblique images were also obtained in several cases (Fig 3) once the proximal coronary artery had been identified.
DISCUSSION We have
demonstrated
the feasibility
of breath-hold MR angiography of the coronary arteries in healthy subjects with use of a standard MR imaging system and body coil. The method proved robust, allowing portions of the proxi-
mal coronary
arteries
to be seen
in all
subjects. The method allowed for acquisition of overlapping 4-mm-thick sections within a brief period of 10-20 mmutes. A major challenge for coronary-flow imaging is to obtain sufficient signal-tonoise ratio to support high-resolution imaging within a body coil. Most highresolution MR angiography has been performed with use of specialized head or surface coils that give higher sensitivity than a body coil. Use of an incremental flip angle has been proposed previously to improve signal-to-noise ratio in fast gradient-echo imaging (7). As applied to flow imaging with our approach, use of an incremental flip angle series permitted application of a large flip angle at the central phase-encoding step without much saturation of inflowing spins, which greatly increased signal-to-noise ratio and flow contrast. This was not the case with use of a constant large flip angle, although centric reordering of the phase-encoding steps might improve the results. Use of an
642
#{149} Radiology
Figure
1.
tamed image, arrow),
with with left
arrow).
Four-millimeter-thick,
breath-hold axial MR images of a healthy subject a 270 x 270-mm field of view, a 12-second imaging
a 128 x 256 matrix, 16 segments of eight phase-encoding anterior descending artery (white
(b) Right
coronary
artery
were
ob-
time per steps each. (a) Left main artery (thick black arrow), and left circumflex artery (thin black
(arrow).
in all subjects.
heart rates per minute,
time right
b.
a.
implemented.
incremental flip angle and k-space segmentatiom has the potential to cause loss of spatial resolution as a result of distortions in the point-spread function. Nonetheless, the resolution was adequate for showing the major coronary arteries. Images obtained with use of this technique were free of ghost artifacts and of blurring from respiratory motion. Substantial improvements in image quality can be expected with faster data acquisition (eg, with use of more efficient pulse sequences and faster gradients). A problem that is less easily overcome is variation in cardiac position from one heartbeat to the next, which causes blurring (10). Potentially, performance of a single-shot technique, such as echo-planar imaging, could overcome this problem, but further development is needed before adequate spatial resolution and signal-to-noise ratio can be obtained by this means. Recently, interleaved spiral scanning, a variation on echo-planar imaging, has shown promise for highresolution imaging of the heart (11). Because interleaved spiral scanning results in fewer saturation effects than gradient-echo methods and is potentially faster, it might prove superior for coronary artery imaging, and further study is needed. Several other techniques have been used to image the coronary arteries. Standard electrocardiography-gated, spin-echo, and cine images have only occasionally depicted portions of the coronary arteries, and these images are not adequate for detailed evaluation (12,13). Three-dimensional acquisition methods also have potential for imaging of the coronary arteries (14). Both the two- and three-dimensional approaches have potential advantages and disad-
vantages. Three-dimensional acquisitions permit use of thinner sections and shorter echo time, thereby minimizing partial volume averaging and flow-related dephasing. With three-dimensional approaches, saturation effects are more severe, resulting in worse flow contrast. The imaging time necessary for three-dimensional acquisitions of more than 1 minute with cardiac gating precludes use of breath holding; therefore, significant blurring from respiratory motion may occur. Further work is needed to determine the relative benefits of the two- and three-dimensional approaches, but one might expect that the methods will be complementary, as they are for MR angiography of the head and neck. Subtraction methods also have potential value for coronary artery imaging. One approach has achieved good background suppression by subtraction of a pair of images, one of which has a spatially localized inversion pulse applied to the aortic root to tag the coronary inflow (15). A major advantage of this method is that a projection image of the coronary arteries is obtained. However, despite initial enthusiasm (16), clinical results with projection angiography have proved disappointing when the technique is applied to other small yessels, such as the carotid arteries. Potential drawbacks to use of the method for coronary artery imaging are the need for excessively long breath-holding periods and a surface coil to achieve adequate signal-to-noise ratio and the possibility of misregistration of small vessels resulting in imperfect subtraction. We have demonstrated the feasibility of performance of breath-hold MR angiography for rapid imaging of the corDecember
1991
onary minimal
arteries. The technique operator intervention
be initiated
with
requires and can
use of a standard
MR
imaging system and body coil. Future efforts will be directed toward achievement of further improvements in spatial resolution and signal-to-noise ratio, as
well
as use of shorter
echo
data
acquisition,
flow
tion.
With
method
sis.
further could
in patients
and
development,
have
with
time,
clinical
coronary
faster
quantifica-
the usefulness
artery
steno-
U
References 1.
Ruggieri PM, Laub GA, Masaryk TJ, Modic MT. Intracranial circulation: pulse-sequence considerations in three-dimensional (volume) MR angiography. Radiology 1989; 171:785-
2.
Masaryk
TJ, Modic
MT.
Ruggieri
PM,
et al.
Three-dimensional (volume) gradient-echo imaging of the carotid bifurcation: preliminary clinical experience. Radiology 1989; 171 :8013.
4. 5.
6.
7.
Kent KC, Edelman RR, Kim D, Steinman TI, Porter DH, Skiliman JJ. Magnetic resonance imaging: a reliable test for the evaluation of proximal atherosclerotic renal arterial stenosis. J Vasc Surg 1991; 13:311-318. Berne RM, Levy MN. Cardiovascular physiology. 5th ed. St Louis: Mosby, 1986; 200. Burstein
Society
8.
d.
C.
Figure left
main
cumflex
2.
Series
of breath-hold
coronary
artery
(arrow).
coronary
artery
(arrow).
axial MR images (b) Left
anterior
show
left coronary
descending
artery
circulation. (arrow).
(a) Ostium (c, d) Left
cir-
of 9.
10.
11. 12.
13. 14.
15.
D.
MR
imaging
of coronary
artery
flow in isolated and in vivo hearts. JMRI 1991; 1:337-346. Atkinson DJ, Edelman RR. Cineangiography of the heart in a single breath hold with a segmented turboFLASH sequence. Radiology 1991; 178:357-360. Stehling MK. Optimised (incremented) rfangle gradient echo imaging: ORANGE (abstr). In: Book of abstracts: Society of Magnetic Resonance in Medicine, 1990. Berkeley, Calif: of Magnetic
Resonance
in Medicine,
1990; 459. Edelman RR, Chien D, Atkinson DJ, Sandstrom J. Fast time-of-flight MR angiography with improved background suppression. Radiology 1991; 179:867-870. Crawley AP, Wood ML, Henkelman RM. Elimination of transverse coherences in FLASH MRI. Magn Reson Med 1988; 8:248Pearlman JD, Weisskoff RM, Hunter MS. Cohen MS. Brady TJ. Cardiac variance images from single-shot MR imaging (abstr). JMRI 1991; 1:181. Macovski A, Meyer CH, Noll DC. Highspeed imaging with time-varying gradients (abstr). JMRI 1991; 1:138. Paulin 5, von Schulthess GK, Fossel E, Krayenbuehl HP. MR imagine of the aortic root and proximal coronary artenes. AJR 1987; 148:665670. Alfidi RJ, Masaryk TJ, Haacke EM, et al. MR angiography of peripheral, carotid, and coronary arteries. AIR 1987; 149:1097-1109. Paschal CB, Haacke EM, Adler LP, et al. High resolution 3D cardiac imaging (abstr). In: Book of abstracts: Society ofMagnetic Resonance in Medicine 1990. Berkeley, Calif: Sodety of Magnetic Resonance in Medicine, 1990; 278. Wang SJ, Hu BS, Macovski A, Nishimura DC. Coronary angiography using fast selective inversion recovery. Magn Reson Med 1991; 18: 417-423.
16.
Figure 3. Oblique views in another healthy subject were obtained with 15-second time per image. (a) View through axis of the left main coronary artery (arrow) and orthogonal to view in a show cross section of left main artery (arrow).
Volume
181
#{149} Number
3
Wedeen VJ, Meuli RA, Edelman RR, et al. Projective imaging of pulsatile flow with magnetic resonance. Science 1985; 230:946-948.
imaging
(b) view
Railinltwv
M
#{149}