Billary
Evaluation
Obstruction:
with
Three-dimensional
MR Cholanloraphy’ Kouji
useful
Mutsuo
ship
Morimoto, MD Shimoi Toyomi Shirakawa Yoshiko Aoki, MD Soomi
Choi,
Yoshiaki
Miyata, Hara, MD
Kazuo
in establishment between
MD
good
MD
age
with
comparable
correlation
performed
0-21
days
later
terms: stenosis
76.365 mensional
Magnetic
resonance
1992;
183:578-580
Radiology
N
Bile ducts, MR, 76.1214 or obstruction, 76.288,
#{149}
ONINVASIVE
been
(MR),
Bile
#{149}
three-di-
has by the
methods, (endoscopic
resonance
and
bile
ducts
evaluate
such cause
In for obmag-
(MR)
projection
(MR
cholangiograms)
the clinical
images
and
effectiveness
of
Methods
Our technique for 3D MR cholangiography was developed and evaluated by a standard 1.5-T clinical wholeimaging system with a 10-mT/rn
gradient
capability
1.0-
msec minimum rise time (Magnetorn; Siemens, Erlangen, Germany). Signal
reversed with
but or
percutaneous) is often then required to determine the cause and exact site of biliary obstruction prior to surgery. CT and US provide tomographic images; direct cholangiography can demonstrate a whole biliary system as a projection image in different planes, which is
version
of the fast imaging
steady-state
nique.
With
precession
(FISP)
this sequence,
to generate
tech-
it is possible
steady-state
free
precession
(SSFP) signals with certain parameters. The SSFP signal has two contributions: a free induction decay component occurring
at the
interval
and
“SSFP
start
an echo
echo”)
at the
of each
repetition
component end,
just
(the as the
next
radio-frequency (RF) pulse is applied. The CE-FAST sequence is used to detect the SSFP echo signals. This sequence is very sensitive to physiologic factors such
as respiratory
and
cardiac
motion.
To avoid signal loss caused by respiratory motion, breath-hold imaging is used. From
the Department
National Hospital, Osaka 540, Japan
K.H.) and Siemens 1991
of Radiology,
2-1-14 Hoenzaka, (KM., MS., Y.A.,
Asahi,
Tokyo
Osaka
the
RSNA scientific assembly. Received September 9, 1991; revision requested October 9; revision received November 14; accepted November 15. Supported in part by a grant-in-aid for cancer research from the Ministry of Health and Welfare, Japan. Address reprint requests to
KM. c
RSNA,
578
#{149}
1992
Radiology
sequence
cannot
detect
blood
flowing at over 1 mm/sec (2). Under these conditions, it provides a heavily T2-weighted image, and only the fluid
Chuo-Ku, S.C., Y.M.,
(T.S.). From
The
in the played
static state, such as bile, is disas signal of extremely high in-
tensity. The coronal imaging, and technique
was
plane was a 2D and/or used.
selected for 3D imaging
Variables
imaging were as follows: tion time [TR] msec/echo
of 2D
17/7 (repetitime [TE]
signal
aver-
and
with
a 350
256
x
x
field of view. Imaging was performed during a breath hold of 16 seconds. Variables of 3D imaging were as follows: 17/7, 90#{176} flip angle, one signal average, 32-mm slab thickness, 4-mm partition thickness, and 128 x 256 acquisition matrix with 350 x 350-mm field of view. Slabs were acquired dura breath
sides
of
hold
of 20
effect
the
slab
seconds.
was
with
seen
3D
on both
imaging
per-
formed under these conditions. These artifacts were eliminated from the 3D data set, and slabs were overlapped to avoid any resultant data gap. 3 D image reconstruction-The images were processed by using a maximumintensity
projection
(3,4).
and
three
thickness,
matrix
An aliasing
images in determination of the and site of biliary obstruction.
Materials
angle,
section
acquisition
ing
3D fast imaging technique with a flowsensitive gradient-echo pulse sequence (contrast-enhanced Fourier-acquired steady-state technique [CE-FASTJ imaging) (1). This sequence is an exact time-
development of computed tomography (CT) and ultrasound (US). Bile duct dilatation and obstruction sites can be well these
netic of the
perfor-
methods (3D)
6-mm
350-mm
detection was provided by a circularly polarized body coil (Siemens). Data acquisition-The method is based on a two-dimensional (2D) and
76.363,
biliary imaging markedly advanced
demonstrated with direct cholangiography
256
cholangio-
later
90#{176} flip
ages,
im-
cholangiography.
this article, we describe taming three-dimensional
maximum
was
without
of direct
using body
between
observed. Index ducts,
a direct
msec),
and
site. If a projection
be determined
findings at 3D MR cholangiography and percutaneous transhepatic biliary drainage
branches
gram could be obtained noninvasively, the precise location of the obstruction and its relation to the biliary trees could
Three-dimensional (3D) magnetic resonance (MR) projection imaging was evaluated as a noninvasive alternative to direct cholangiography in 12 patients with malignancy-related obstructive jaundice. The 3D images of the bile ducts were formed by subjecting consecutive coronal MR images obtained with a fast imaging method to a maximum-intensity projection algorithm. Dilatation and obstruction of the biliary system were well documented in all and
of the relation-
biliary
the obstruction
mance
cases,
the
With
this
(MIP) method,
algorithm the
images
are
viewed at a user-selected projection angle, and the pixels of highest intensity are identified. These pixels are assumed to represent bile ducts, and their signal intensities are assigned to pixels in the MR cholangiogram.
Two about
recalculated 10#{176} apart
images
for views
could
be observed stereoscopically. The MR cholangiogram was constructed in multiple projections 3#{176}-5#{176} apart and was then displayed in cine mode. These methods allow clear demonstration of the 3D configuration of the bile duct. Clinical application-Twelve patients (10 men, two women) with obstructive jaundice underwent MR cholangiograpity followed by percutaneous transhepatic biliary drainage (PTBD). Dilatation of bile ducts had been previously demonstrated with CT and/or US. Patient age range was 40-85 years (mean, 67.8 years), and patients had pancreatic cancer (i = 3), bile duct cancer (i = 7), or gastric
cancer
(i
benign cluded
causes in the
of obstruction were instudy. The interval be-
tween
=
2).
MR cholangiography
No
patients
and
with
PTBD
(mean, 6 days). At the time of MR cholangiography, the serum bilirubin concentration was 5.3-41.0 mg/dL (91-701 .mol/L), with a mean of was
0-21
15.2
mg/dL
days
(260
p.mol/L).
Results In all cases, dilatation and obstruction of the bile ducts were clearly demonstrated. The hilum of the liver (n = 7) and extrahepatic bile duct (i = 4) or both (n = 1) were the documented sites of obstruction. The findings at MR
May
1992
b.
a. Figure
1.
(a) MIP MR cholangiogram
(full data
C.
set) obtained
in a 70-year-old
man
with
recurrent
gastric
cancer
and
related
obstructive
jaun-
dice shows dilatation of intrahepatic bile ducts and stricture at the hilum of the liver. The extrahepatic duct was also dilated, and obstruction at the end of the conimon bile duct was suspected. (b) MIP image (selected data set) reveals dilated pancreatic duct. (c) PTBD and internal drainage were performed 3 days after MR cholangiography. This direct cholangiogram shows good correlation with the MR cholangiogram (a). Advanced tumor invasion into the hilum of the liver and the duodenum of the peripancreatic area was confirmed at surgery.
cholangiography
were
exactly
corre-
lated with those at PTBD. Dilated creatic ducts also were delineated two cases (pancreatic head cancer recurrence of gastric cancer) (Fig Bile
ducts
strated
or
gallbladder
by means
phy
were
demon-
cholangiogra-
demonstrated in five
cholangiography 2a,
not
of direct
panin and lb).
with
MR
cases
(Figs
la,
3a).
Images
obtained
showed
with
higher
those
obtained
with
With
CE-FAST
data
aging
time
half
the
can
the
2D
data
set. total
shortened
taken
set
im-
to about
with
the
2D
Fourier
a.
method.
With
a stereoscopic
display,
bile
3D data did
than
sampling,
be
time
transform
the
resolution
the
ducts
view
spatial
and
cine
relationship
could
of
be easily
recognized.
technique
is one
b.
Figure
mode
2. (a) MR cholangiogram obtained in a 66-year-old man with bile duct cancer. Dilatation of the intrahepatic bile ducts and gallbladder and obstruction at the common hepatic bile duct are demonstrated. (b) PTBD was performed 3 days after MR cholangiography. The sites of obstruction on the MR cholangiogram are also evident on this direct cholangiogram, but the gallbladder could not be opacified at direct cholangiography.
the
Discussion The
CE-FAST
MR imaging In
the
methods
SSFP
sequence,
RF pulses
of the
plied
with
free
induction
pulse
next
sequence, echo)
With
pulse
the
assumption
acts
second-to-last
RF
would
equal
two
which
is longer
resulting
from
sequence
(5).
existing not
than the To
This
T2-weighted
acquisition
avoid
long
(the
the
true
TE can
creased TE
with
TR
provide
cacy
Se-
a
Number
2
of
suitable bile
respect
longer
the
is displayed signal.
as
These
for
selective
duct
system.
to TR
the
was
contrast
of a 2D
evaluation to the data fast
#{149}
sensi-
inition
of
1 mm/
2D
read
gra-
extremely visualiza-
imaging
fast
associated
PTBD
paramewith
the
imaging
bile
method
showed
duct
tree
method,
and
than
did
the
because
a
in
drainage are
better
treatment
and
(7).
these branch
For
of
must
insertion cially
effective
interventions, be
of the when
the
inter-
patients
with
jaundice an opera-
performance
the
of
optimal
selected
as the
drainage
tube.
obstruction
patic or at the hilum length of the segment
bili-
important
malignancy-related obstructive that cannot be resolved with tion
def-
internal
endoprosthesis
ventions
in
better
ary
de-
of bile ducts (6). With respect sampling method, the 3D method
biliary
imaging
technique.
imaging
between
of the fast
contrast between bile duct and surrounding soft tissue was obtained with the 3D fast imaging method. Also, the total imaging time with the 3D sequence is shorter than that with the 2D sequence. Thus, the 3D fast imaging method appears to be superior for this
characteris-
and surrounding tissues on the constructed MR cholangiogram, although flip angle of greater than 40#{176} did not have much effect in our limited research. Wallner et al reported the effi-
TE,
in a short
direction
state
of the
ters,
TE
>
are With
RF
FISP
TE
image
tics
of the
CE-FAST
static
tion
last
true
the
high-intensity
(1).
confusion the
the
end
indicated
time.
183
at the
In a CE-
time-reversed
with
heavily
each RF RF pulse
the the
order
dient (2). Under such conditions, signal from blood flow is eliminated almost completely. On the other hand, fluid in
TR minus
nomenclature,
adopted
quence.
Volume
pulse,
is very
the
sec)
signal
times
(on
ap-
images
the
along
sequence
flow
The
that
to refocus
slow
are
component
to form
CE-FAST
to
TR.
signal
this echo
The tive
short
interval.
is used
(1).
angle
following the next
by
repetition
SSFP
was
flip and
decay
FAST
the
same
a constant
a spin-echo-like
of the
of the
on SSFP
phase-coherent
is refocused
to form
based
biliary site
for
Espe-
is intrahe-
of the liver, of drainage
Radiology
the tube #{149} 579
inferior to that with direct cholangiography, both maneuvers result in a similar standard of images. The information obtainable with either method is of similar good quality, but the noninvasive procedure is less traumatic for the patient. #{149}
References 1.
2.
Cyngell ML. The application of steadystate free precession in rapid 2DFT NMR imaging: fast and CE-fast sequences. Magn Reson Imaging 1988; 6:415-419. Patz 5, Hawkes RC. The application of steady-state free precession to the study of very slow fluid flow. Magn Reson Med 1986; 3:140-145.
3.
a.
b.
Figure
3.
shows
marked
(a) MR cholangiogram dilatation
(b) PTBD and internal after MR cholangiography. tion
with
MR
obtained
of the
but
bile
ment) branch
is limited, and a longer biliary must be selected for effective
drainage.
(effective
the
in the
view great
drainage
Three-dimensional
cholangiography,
bile
with
left
branches
seg-
MR stereoscopic
and/or cine mode display, is of help in selection of the appropri-
ate approach and effective execution tions.
For patients dice, noninvasive projection image
limited method
insertion of these
with
duct
man
and
with
obstruction
point for interven-
obstructive
jaun-
methods to obtain of bile ducts have
to scintigraphy is not effective
(8,9). This in patients
the
could
#{149} Radiology
liver.
performed 1 day shows exact correla-
be opacified.
of tumor
invasion,
picted clearly. MR cholangiography, method
for constructing
4.
5.
could
6.
7.
8.
be de9.
as a noninvathe
tion image of bile ducts, has great tential in the diagnosis of biliary disease. Although spatial with 3D MR cholangiography
580
cancer
of the
level. In addition, nonopacified areas on the direct cholangiogram, such as sequestrated bile duct or gallbladder as a
sive
with
were
gastric hilum
severe jaundice and does not provide sufficient image quality for planning biliary drainage and surgery. On the other hand, in our preliminary research, MR cholangiography was able to delineate the dilated biliary systern irrespective of the serum bilirubin
result the been
not
advanced at the
drainage of the right branches of the bile duct In the right branches, direct cholangiography
cholangiography,
duct
in a 40-year-old
intrahepatic
projec-
Anderson CM, Saloner D, Tsuruda JS, Shapeero LC, Lee RE. Artifacts in maximumintensity-projection display of MR angiograms. AJR 1990; 154:623-629. Edelman RR, Mattle HP, Atkinson DJ, Hoogewoud HM. MR angiography. AJR 1990; 154:937-946. Lee SY, Cho ZH. Fast SSFP gradient echo sequence for simultaneous acquisitions of FID and echo signals. Magn Reson Med 1988; 8:142-150. Wallner BK, Schumacher KA, Weidenmaier w, Friedrich JM. Dilated biliary tract: evaluation with MR cholangiography with a T2-weighted CE-FAST sequence. Radiology 1991; 181:805-808. Mueller PR. Interventional radiology of the biliary tract: a decade of progress. Radiology 1988; 168:328-330. Lee AW, Ram MD, Shih WJ, Murphy K. Technetium-99m BIDA biliary scintigraphy in the evaluation of the jaundiced patient. Nucl Med 1986; 27:1407-1412. Rosenthall L. Cholescintigraphy in the presence ofjaundice utilizing Tc-IDA. Semin NucI Med 1982; 12:53-63.
po-
resolution is slightly
May 1992