Donald G. Mitchell, Keith Haidet, MD
MD #{149}Simon Vinitski, PhD Matthew D. Rifkin, MD
#{149} Feroze
B. Mohamed,
BS
Comparison ofKaopectate for Negative and Positive Contrast at MR Imaging’
T
with Enteric
evolution of clinical magnetic resonance (MR) imaging of the upper abdomen has been impeded by the tack of a suitable orally administened contrast agent. It is often difficult to distinguish between normal and pathologic bowel tissue; a pathologic condition intrinsic to the bowel can rarely be imaged adequately at
tional
MR.
containing
from
0%
to 100% attapulgite or barium preparations were compared by using spinecho and spoiled-gradient-echo techniques
to estimate
Ti,
T2,
and
T2*
relaxation times. In vivo spin-echo and gradient-echo images were obtamed of 10 healthy volunteers and of five patients with pancreatic disease after oral administration of 12-16 oz of attapulgite preparation. Ti and T2 relaxation times obtained with use of attapulgite were slightly lower than those obtained with kaolin-pectate at both 50% and 25% concentrations.
Compared
with
barium,
attapulgite had shorter Ti, T2, and 12* relaxation times, and concentration-dependent relaxivity was greater by
nearly
an order
subjects
were
able
of magnitude. to drink
All
the
atta-
pulgite, and the only complaints were of bad taste and mildly loose stools. Index
terms:
Contrast
Abdomen,
media,
tines,
MR
(MR),
contrast
Radiology
studies,
MR
comparative 70.1214
studies, studies
Magnetic
70.1214 #{149} Intesresonance
enhancement 1991;
19107.
181:475-480
Received
PhD
replaced computed tomography (CT) as the primary screening modality for the abdomen (1-3). In addition, MR imaging is rarely performed to enable detection, characterization, or staging of pancreatic disease, for which CT is often performed. Dilute preparations of barium or iodinated contrast agents, routinely administered orally prior to CT imaging of the abdomen, greatly increase the ability of CT to depict and enable characterization of extrahepatic pathologic conditions (4). These agents are widely used because of their high contrast relative to other tissues, as well as their safety and tow cost. These important attributes should be considered in the search for an orally administered contrast agent for use at MR imaging. The choice between increasing the signal intensity of intestinal contents (“positive contrast”) or suppressing it (“ negative contrast”) depends on the pulse sequence used. On T2-weighted images,
whether
HE
Therefore,
negative
MR
imaging
contrast
has
not
is prefera-
ble, since this decreases artifact and prevents confusion between normal bowel tissue and pathologic bowel tissue that has high signal intensity
, From the Department of Radiotogy, Thomas Jefferson University Hospital and Jefferson Medical College, 11th and Sansom Sts, PhiladelPA
MD,
wall. In addition, suppressing the signat of bowel, as welt as that of fat, neduces artifact and optimizes dynamic range, impnoving depiction of Ti contrast (9,10). The diamagnetic properties of clay compounds such as Kaopectate (kaotin-pectate; Upjohn, Kalamazoo, Mich) have been described by Listinsky and Bryant (5). The low cost and established safety of this class of agents are significant advantages over other orally administered contrast agents proposed for use at MR imaging, such as magnetic particles (8), paramagnetic solutions (11), and perfluorocarbon fluids (7). By binding water, clay compounds shorten Ti and 12 relaxation times, resulting in low signal intensity in the bowel on conventional spin-echo MR images. If echo time (TE) is short enough, howeven, 12-shortening effects may be minimized, “unmasking” Ti -decreasing properties. The goals of this study were (a) to determine whether the current formutation of Kaopectate (attapulgite) is as effective as the old formulation of Kaopectate (kaolin-pectate), which was evaluated by Listinsky and Bryant (5); (b) to compare the Ti- and 12-decreasing properties of attapulgite with those of barium, another low-cost contrast agent for use at MR imaging (6); and (c) to determine
(5-8).
phia,
F. Mammone,
Barium
In an evaluation of safe, inexpensive, orally administered contrast agents for use at magnetic resonance imaging, different concentrations of the clay agents kaolin-pectate and attapulgite were compared in an in vitro phantom by performing spin-echo imaging with variable echo times to estimate 12 relaxation time. Addiphantoms
#{149} Joseph
#{149}
March
27,
1991;
revi-
sion requested April 30; revision received June 10; accepted June 17. Address reprint requests to D.G.M. RSNA, 1991
On
Ti-weighted
images,
how-
ever, normal bowel tissue with tow signal intensity may mimic pathologic tissue with tow signal intensity. Therefore, it may be preferable to increase the signal intensity of intestinal contents on conventional Tiweighted images. With fat suppression, however, negative contrast may produce a double contrast” effect, improving depiction of the bowel “
attapulgite
enha
nces
lumen effectively on spin-echo ultrashort-TE, gradient-echo ages. MATERIALS Kaopectate
AND was
bowel
and MR im-
METHODS
purchased
in a super-
market for approximately $5.00 per i2-oz bottle. Alt MR examinations were performed with a 1.5-1 unit (Signa; GE Medicat
Systems,
Abbreviations: tion
Milwaukee).
TE = echo
Three
time,
phantoms,
TR = repeti-
time.
475
each
consisting
were
imaged
and
of a vertical with
transmit
agitated
use
of vials, receive
attenuation.
Alt
approximately
to the
onset
vials
tate (kaolin, mg/mm )
vials
The
prior
first
mg/mm3,
scnibed
Kaopec-
and
and new-formula
pectate,
above.
4.3
Kaopectate
ing
a three-point
was
calculated
(attapulgite, 37.5 mg/mm) in 100% , 50%, and 25% concentrations. These phantoms were imaged in the sagittat plane by using
(1,500/9,
a single-section, repetition time
above. The the relaxation
20 msec
spin-echo (TR) of 400
(TRITE
to estimate each vial
wt/wt; volume
This
tions.
contained
and
E-Z-Em, at both
barium
Westbury, 100% and
phantom
imaged
to estimate
and
1,600
50,
msec.
with 100,
by using
ages.
partial-echo sampled
was measured background.
=
Since much
p
(S)
in spin-echo less
than
(in which approximate
T2*
deviation imaging,
of we
(1)
experiments
TR,
we
used
Equation
(I)
12)
to were
to provide a comparison among of these solutions at MR imagthan to provide a definitive
measure
of the
actual
relaxation
the
comparison
barium
water
wt/wt
found
In
to be
precipitated
containing 60%
E-Z-Em)
taming preparations 40%, 60%, 80%,
#{149} Radiology
for within
a premixed
(Novapaque;
for
and
dilution.
wt/wt was
this
The
was
experiment,
20%
or more
minutes.
There-
preparation
used.
White in these
dependence 1/12, and
correlated
a. of
1/12*
with
Fisher
the using Softsusvis-
Scientific,
the exact shear measurements,
tow because experiments.
Human subjects investigation
(2,500/50, were
of the were approved
prior
after bottles)
shear
stress
imaged as part by our institu-
of C.
Conventional TIand T2-weighted
100) images
obtained
low
of 10 volunteers to and
approximately
administration of attapulgite.
with
Vials
of 0%, 10%, 20%, 100% concentrations
quences bottle
after ingestion of attapulgite.
of 16 oz Higher
pancreatic
The
b.
rate is it is rel-
of the contrast agent, use of which may have improved imaging, were not given because of concern regarding patient compliance. Subjects were polled for any adverse effects within 48 hours
for time-of-flight tionat TI- and ages
and
of barium
unsuitable
all dilutions
of
of concentrations;
used
was
using
Figure
1.
tamed
with
In vitro TR
spin-echo
of 2,000
images
msec
ob-
compare
kaolin-
pectate (old-formulation Kaopectate) and attapulgite (new-formulation Kaopectate) at 25%, 50%, and 100% concentrations with tap water. W = water, A = attaputgite, K = kaolin-pectate, /4 = 25% concentration, /2 = 50%
concentration.
Both
formulations
are
msec, msec.
= 40
(b) TE
100%
signal
msec,
concentration
voids.
(a)
and
(c) TE
TE
=
20
= 80
cancer imdif-
of a single 12-oz first gradient-
quence and in the axial plane at the end of the examination. Single-section, gradient-echo (27/7.4) images with a 20#{176} flip angle were obtained of each subject
to expand
attapulgite
preparation
water
476
imaged
range
was
since fore,
was between
at a wider
distilled
85%
times.
the validity of measurements time that use truncated radiopulses has not, to our knowl-
edge, been proved. A third phantom
a two-point
echo sequence consisted of 12 radio-frequency-spoiled (101/2.3) images acquired within a 14-second suspended respiration. These images were obtained in the coronal plane as an initial localization se-
TE is
obtained properties ing rather particular, relaxation frequency
using
and one with pancreatitis underwent aging with use of spin-echo and two ferent breath-hold gradient-echo se-
‘#{149}
is substituted with T2. The measurements
and
after ingestion. Four patients
is
I2(1
t
us-
time
quantities
image.
for a 90#{176} radio-frequency intensity
cal-
was estimated by technique described
(Canon-Fenske;
30 minutes (two 8-oz
dephasing of transby radio-frequency
Thus,
S
im-
measurements noise, which
as the standard In gradient-echo
signal
of
by using full echo
the
20-msec
intensity to system
assumed complete verse magnetization
spoiling.
while
the
for
Therefore, these were normalized
pulse,
acquisitions 20)
by
tional review board. weighted (400/12-14)
was
achieved
sampling,
an
800,
(4,000/12,
was
was obtained TR by
double-spin-echo
images
plotted
atively in these
msec
relaxation
separate
12 msec
400,
time
images variable
concentration rates 1/Ti,
shear rate. unknown
TI reof 2.1
200,
T2
spin-echo of
TE
TE
150,
Finally,
single-section,
was
T2* relaxation.
estimated
of 25,
estimated
a
pulses. TR was 100 msec, varied (2.1, 3, 4, 6, 8, 10, and
was TRs
with
from
ar-
Pittsburgh) at 20#{176}C.For strictly quantitative measurements, the use of a capillary viscometer assumes that the fluid obeys Poiseuille law. This assumption is not entirety valid for suspensions in which the apparent viscosity varies according to the
NY) in an equal 50% concentra-
was
relaxation
fit. T2 relaxation
18)
cometer
of 85%
use of a single section and Prototype short-TE softthat employed truncated
radio-frequency while TE was
and
not
vials
was
barium,
fit of the linear regression curve by Cricket Graph 1.3 software (Cricket ware, Malvern, Pa). Kinematic viscosity of the various pensions was measured in a capillary
gradient-echo
technique, with a 90#{176} flip angle. ware was used
15 msec)
40 of
(HZ-HD,
radio-frequency-spoiled,
taxation
2,000/20,
time our experishorter TE was
phantom
attapulgite
and
TI
fit. T2* relaxation the gradient-echo
with TE of
T2 relaxation times use of region-of-interest
At the performed,
available. The second both
400/20)
=
TI and through
measurements. ment was
technique, msec and
and
rows, were imaged technique de-
culated from spin-echo with TE of 9 msec and
phantom
of original-formula
194
of attapulgite
ranged in two separate with the single-section
were
15 minutes
of imaging.
contained
by volume
row
of constant
con-
were
flow imaging. Conven12-weighted spin-echo
obtained
for
the
normal
im-
volun-
teers, as well as TI-weighted images in which fat was suppressed with a combination of selective saturation and opposed-phase techniques (9,12). The signal intensity of the mens was noted,
gastric and duodenat tuas was visibility of the
head, body, and tail of the pancreas. Because of the limited time between ingestion of the contrast material and imaging and because of the small volume of contrast agent remainder evaluated.
that was administered, of the small bowel
was
the not
RESULTS Alt curve tions
fits
resulted
mated
Ti and
tamed
with
tions
than
Table
1). The of too
the
relatively
in
this
the
The
.99.
>
12 relaxation and
slightly
pulgite
were
relaxation
r
50%
were
times
for in
for
25%
shorter
for
to tong
atta(Fig
1;
12 relaxation
full-strength
experiment.
concentra-
kaotin-pectate
Ti and
short
estimaestitimes ob-
formulations
be
estimated IRs
and
Because
with TEs
used
of limited November
1991
a.
b.
Figure
2.
In vitro images contain 100%, 80%, 60%, nat intensity than distilled (1,000/6) image. All tubes signal ing
intensity
than
60%-I00%
C.
compare attapulgite and barium (60% wt/wt). lop 40%, 20%, 10%, and 0% concentrations. (a) Ti-weighted, water. Tubes containing 20%-60% attapulgite have containing barium have higher signal intensity than
water.
(c)
attapulgite
(1,500/18)
Spin-echo
have
lower
signal
image.
intensity
Alt tubes
than
with
row
= barium,
bottom
gradient-echo the highest signal water, while tubes
barium
have
higher
row (100/2.2) intensity. containing
signal
= attapulgite.
From left to right, image. Alt samples have higher (b) T2*weighted, gradient-echo 60%-I00% attaputgite have
intensity
than
water,
while
tubes
tubes sigtower
contain-
water.
20
200 :;
150
100
50 t006’08’01#{224}0 20
,
40
a. Figure 60%
60
80
100
Concentration
%
%
3. wt/wt
Relaxivity
plots
show
the
concentration
dependence
of (a) 1/Ti,
(b) l/T2,
(c) 1/T2*
for 50% dilution (4.0 msec vs and 8.3 msec, respectively, for 12* relaxation time and 10.2 msec vs 11.4
attaputgite
varied
centration,
but
and
tapulgite
29.1
msec,
of
two Even
data points, the accuracy of these relaxation measurements is limited, but it is sufficient to establish that attapulgite is at least as effective an MR contrast
as kaolin-pectate. had shorter 12,
agent
Attapulgite
Ti relaxation
times
12*,
barium Compared with either attapulgite or barium alone, the mixture of the two preparations had shorter 12* relaxation time (i.8 msec vs 2.4 and 3.6 msec, respectively) and 12 relaxation time (4.3 msec vs 5.5 and 6.5 msec, concentrations
respectively).
Volume
for attapulgite
similar
the
all
rates
4.9
cause
at
relaxation
(U)
and
barium
(#{149}).
respectively,
for
12
relaxation time). The 50% concentration of attapulgite had approximately the same catculated 12 in the first two experiments (11.0 msec vs 11.4 msec), but the calculated Ti was different (118 msec vs 78 msec). This may be be-
and
Concentration
C.
LI
This
181
than
(Figs
#{149} Number
relationship
2
2, 3).
was
tated ium
technical
between
experiments. though the vials were agiprior to imaging, the 50% bar-
mixture
Repeated 85%
differences
imaging
separated
noticeably.
experiments
wt/wt
barium
confirmed settled
that
within
minutes when diluted to 20% or more water by volume. No gross evidence of settling of any of the other solutions
was
noted
at
imaging
or
an
order
dilutions cause The
1/12*
relaxation
rates
i/T2, for
used
and
at-
be calculated
wt/wt
60%
barium
wt/wt
greater
than
at high
concentrations
too
of
was
barium,
especially
(Table
drank
for
be-
rapidly.
attapulgite
that
All subjects
at or
barium,
precipitated of
by for
concentration,
of 85% they
for
viscometers
the
2).
attapulgite,
but all found it unpleasant. objected to both the taste
Subjects and consistency, and some stated that it became increasingly difficult to drink because of a feeling of gastric fullness. There were no reports of constipation or pain
after
ingestion,
stools In
although
reported
that
were
volunteers, resulted
except
in
an
distention
anterior
of
loose
accompanied
ingestion in
five
mildly
not
ach and duodenum virtual signal void
and
barium
not
60%
viscosity
gite
for the third experiment. The calculated i/Ti,
than
10 volunteers
visual
therefore
or
greater
con-
nearly
greater
capillary
attapulgite
ods
were
the
with
was
3).
could
Viscosity
cramps.
and
magnitude
(Fig
using
inspection within 1 hour of mixing. Dilutions of 60% wt/wt barium nemained in suspension for longer penof time
of
linearly
the slope
by
of
attaput-
of
the
and produced at MR imaging, layer of fluid Radiology
stom-
a
#{149} 477
with intermediate to high signal intensity. Depiction of the pancreas was improved after administration of attapulgite (Fig 4). In four of five patients, the bowel lumen demonstrated a near signal void on images obtained with TE not tess than 7.4 msec. The fifth patient had partial obstruction of the gastric outlet, which resulted in dilution of attapulgite with gastric secretions. In this patient, the bowel lumen demonstrated high signal intensity on alt images obtained with TE of not more than 12 msec, but the signal intensity of bowel was low on images obtained with TE of not less than 50 msec. In all patients, the pancreas, stomach, and small bowel were delineated
clearly
on
all enhanced
a.
b.
images
(Fig 5). Benign and malignant pancreatic tissue could be clearly diffenentiated. The stomach and the duodenum demonstrated the same signal intensity. d.
C.
DISCUSSION Attapulgite
Figure
is a safe,
inexpensive,
and readily available orally administered contrast agent with promising effects on shortening Ti and 12. Retaxation is enhanced even when attaputgite is diluted, as might occur in the presence of intestinal secretions. There the
are
abundant
broad
margin
administration other
of of
clay
The
data
establishing
safety
for
preparations
oral
Li
et
at (6).
because
preparations
appear of
their
to
be
at
unsuitable
tendency
to
precipi-
tate with even mild dilution, presumably because of the limited amount of suspending agents. Paramagnetic (11), supenparamagnetic (8), and perfluorocarbon (7) agents have shown promise in clinical trials, but the production and investigation of these agents may be expensive. For 12-weighted images, in which it is desirable to decrease the signal intensity of the bowel lumen, attapulgite was effective as a contrast agent in alt cases, including in a patient with
gastric
distention
that
was
the
of partial obstruction of the tric outlet. In 12-weighted images tamed with barium as the contrast agent, signal intensity was higher the calcutated 12 and 12* relaxation result
478
#{149} Radiology
were
with
(5).
Barium
times
on
relaxation
wt/wt
of the pancreatic
et
properties of attapulgite appear to be significantly supenion to those of even the most concentnated solutions of barium in use clinically. For our concentration-dependent nelaxivity data, we used a 60% wt/wt preparation of barium, which was reported to be optimal by 85%
piction
consistent
and
attapulgite
4. (a, c) Ti-weighted (400/14) images tamed (a, b) prior to and (c, d) approximately by a volunteer. Except for a thin layer of high tow signal intensity on both TI- and T2-weighted
gasoband
longer, with
at that 12-weighted orally
signal
body
and
findings the
images administered
tail (P).
that
observation
reduction
were of
was
Li
limited
enhanced barium
(6).
Li et at noted that barium provided the best negative contrast on 11weighted images, but water alone also could provide good negative contrast because of its long 11 relaxation time. For Ti-weighted images, the deci-
to increase or decrease the signal intensity of the lumen is tess obvious. Achievement of negative contrast decreases the potential for the appearance of ghost artifact on TI-weighted images, but ghost artifact can be avoided by decreasing TE and imaging time, especially if respiration can be suspended. White we occasionally note significant bowel-related artifact on conventional spin-echo images obtained with TE of 12 msec (probably because of respiratory-induced motion of hypenintense fluid), we have not noted this artifact on breathhold images obtained with TE of 2.3 msec. The high signal intensity of bowel can also potentially obscure vessels on MR angiognams. In our series, the 12* relaxation time of attapulgite was sufficiently short to produce low intraluminal signal intensity on att images obtained with TE of not less than 7 sion
and (b, d) T2-weighted (2,500/100) images ob30 minutes after ingestion of 16 oz of attapulgite signal intensity at the top, the gastric lumen has images after ingestion, which improves de-
msec,
except
struction
the
in of
the
patient
gastric
with
outlet.
ob-
Nega-
tive contrast Ti-weighted
was also beneficial images acquired
on with fat suppression; the bowel watt was depicted as having high signal intensity relative to fat on one side and relative to the contrast agent on the other. White it is possible that intermediate signal intensity might result from the use of certain pulse sequences, this did not occur in our series. Positive contrast can be beneficial on Ti-weighted images because of increased differentiation of the bowel lumen from most other tissues. In all subjects,
bowel
the
lumen
demon-
strated
high signal intensity on images obtained with TE of 2.3 msec because of the short Ti relaxation time of the clay agent. The pancreas was depicted as hypointense to fat and bowel lumen but hypenintense to tumors
or
other
abdominal
viscera.
Although all subjects drank the attapulgite in this series, tack of palatability remains a significant obstacle to routine use of currently available clay-based formulations. The high viscosity
of
attapulgite,
which
is even
greater
than that of barium, contnibutes to the unpleasant experience of ingesting large quantities. Variable patient toterance is already a limitation with MR imaging because of the November
1991
a.
b.
d.
C.
e.
Figure
5.
creatitis. duodenat intensity
(101/2.3) image was obtained with a flip angle of 90#{176}. Attapulgite within the gastric (G) and (D) lumens has high signal intensity, allowing clear depiction of the mural folds of both structures. A pseudocyst (C) with high signal is adjacent to but separate from the lesser curvature of the stomach. The pancreatic head (P) is prominent. Small-bowel contents arrows) also have high signal intensity. Black lines at fat-water interfaces indicate partial volume effects between fat and water, the of which are opposed at TE of 2.3 msec. Large arrow indicates susceptibility artifact from metallic clips in the gallbladder fossa. (b) Coro-
(small
Images (a) Coronal,
obtained after Ti-weighted,
administration gradient-echo
of 12 oz of attapulgite
in a patient
with
pancreatic
pseudocysts
that
occurred
after
acute
pan-
phases nat, gradient-echo (27/7.4) image obtained with a flip angle of 20#{176} corresponds to the image in a. Attapulgite within the gastric (G) and duodenat (D) lumens has low signal intensity. As in a, the walls of each viscus are depicted clearly. C = pseudocyst, P = large pancreatic head. Fluid in the small intestine (arrows) also has low signal intensity. The celiac trunk, superior mesenteric artery, and other vessels are depicted as haying high signal intensity. (c) Axial, fat-suppressed, Ti-weighted (400/i4) image. Attaputgite within the gastric lumen (C) has tow signal intensity, which allows clear depiction of the gastric wall with “double contrast” relative to lumen and fat. The pseudocyst (C), with high signal intensity, can be clearly distinguished from the stomach and from the left adrenal gland (arrow). (d) Axial fat-suppressed image at a tower level reveals a pseudocyst (C), with tow signal intensity, anterior to the left psoas muscle. Attapulgite, with tow signal intensity, facilitates detineation of the duodenal (D) and gastric watts. The superior mesenteric vein (white arrow) and duodenum (D) mark the lateral borders of the enlarged pancreatic head (tong black arrow). Short black arrows indicate the pancreatic duct within the pancreatic head. (e) Axial T2-weighted (2,500/100) image corresponding to the image in d reveals a signal void (white arrow) within the duodenum and stomach. Black arrow mdicates the pancreatic duct. The pseudocyst has high signal intensity.
for long examination times in an enclosed space, accompanied by acoustic noise. We are, therefore, reluctant to add to the possible unpleasantness of MR examinations. In our practice, attapulgite is restricted to use in patients referred primarily for imaging of the pancreas, in which use of orally administered contrast matenat is especially important. Thus far, we have not optimized oral administration of contrast agents, which ne-
comthat have but are more palatable. Some clays, such as bentonite, have especially efficient relaxation properties (5), which might allow them to be used in much tower concentrations than are necessary with attapulgite. Rectally administered clay agents are effective as contrast material because lack of palat-
cessitates
The benefit tered clay-based
necessity
patients
arriving
early
a large volume of contrast nat in divided doses during the preceding the MR examination. drink
Volume
181
#{149} Number
2
to
matehour
Perhaps
pounds similar
ability
other
clay-based
may be developed relaxation properties
is not
a problem
(13).
of using orally adminisor other contrast agents for MR imaging of the abdomen may be greatest for imaging of
the pancreas, which bowel. The pancreas relaxation time, and
is surrounded by has a short Ti its signal inten-
sity
images
on
Ti-weighted
is even
higher than that of the liver (9,10). Pancreatic neoplasms thus have significantly lower signal intensity than normal parenchyma. Motion-artifact correction and fat suppression have been shown to improve depiction of normal and pathologic pancreatic anatomy by unmasking the high tissue contrast available in this region (9,10). As demonstrated herein, use of breath
tered
holding
contrast
and
material
orally
adminis-
at MR imaging Radiology
#{149} 479
may be beneficial. Further evaluation of techniques designed to optimize MR imaging of the pancreas and comparison with other modalities is therefone indicated. U
3.
4.
5.
ACknowledgments: Listinsky, MD, vice; to Esmond
We are grateful to Jay for valuable comments and adMapp, MD, for assistance with barium preparations; to Ken Goodman and Fred Ross for photography; and to Pam Bittle for secretanial
6.
assistance. 7.
References 1.
Barakos J, Goldberg H, Brown JJ, et at. Comparison of computed tomography magnetic
2.
480
resonance
tion of focal Radiol 1990; Chezmar JL, et at. Liver patients with ing. Radiology
#{149} Radiology
hepatic
imaging
lesions.
in the
and
evalua-
Gastrointest
8.
15:93-101.
Rumancik WM, Megibow AJ, and abdominal screening in cancer: CT versus MR imag1988; 168:43-47.
9.
Vassitiades VG, Foley WD, Alarcon J, et at. Hepatic metastases: CT versus MR imaging at 1.5 T (abstr). Radiology 1989; 173(P):175. Garrett PR, Meshkow SL, Pertmutter GS. Oral contrast agents in CT of the abdomen. Radiology 1984; 153:545-546. Listinsky JJ, Bryant RG. Gastrointestinal contrast agents: a diamagnetic approach. Magn Reson Med 1988; 8:285-292. Li KCP, Tart RP, Fitzsimmons JR, et at. Barium sulfate suspension as a negative oral MRI contrast agent: in vitro and human optimization studies. Magn Reson Imaging 1991; 9:141-150. Mattrey RF, Hajek PC, Gylys-Monn VM, et at. Perfiuorochemicals as gastrointestinal contrast agents for MR imaging: prelimi. nary studies in rats and humans. AIR 1987; 148:1259-1263. Hahn PF, Stark DD, Saini S. Lewis JM, Wittenberg J, Ferrucci JT. Ferrite particles for bowel contrast in MR imaging: design issues and feasibility studies. Radiology 1987; 164:37-41. Mitchell DG, Vinitski 5, Saponaro S. et at. The liver and pancreas: improved spin-
10.
11.
12.
echo Ti contrast by shorter TE and fat suppression at 1.5 T. Radiology 1991; 178:6771. Semetka RC, Chew WM, Hricak H, et at. Fat-saturation MR imaging of the upper abdomen. AIR 1990; 155:111-116. Laniado M, Kornmesser W, Hamm B, Clauss W, Weinmann HJ, Felix R. MR imaging of the gastrointestinal tract: value of Gd-DTPA. AJR 1988; 150:817-821. Listerud J, Isaac G, Chan T, Lenkinski RE. Atiphatic/olefinic fat cancellation: optimization and extension (abstr). In: Book of abstracts: Society of Magnetic Resonance in Medicine 1990. Berkeley, Calif: Society of
Magnetic 13.
Resonance
in Medicine,
1990;
590. Neuerburg JM, Klose KC, Bohndorf K, Guenther RW. Kaolin as a negative gastrointestinat contrast agent in MR imaging of colorectat disease (abstr). Radiology 1990; 177(P):325.
November
1991