J Radiol 2007;88:1689-94

© Éditions Françaises de Radiologie, Paris, 2007

Édité par Elsevier Masson SAS. Tous droits réservés original article gastrointestinal

Pineapple juice as a negative oral contrast agent

in magnetic resonance cholangiopancreatography*
L Arrivé (1), C Coudray (1), L Azizi (1), M Lewin (1), C Hoeffel (1), L Monnier-Cholley (1), C Lacombe (1),
S Vautier (2), J Poupon (2) and JM Tubiana (1)

Résumé
Utilisation du jus d’ananas en tant qu’agent de contraste négatif en
cholangiopancréatographie par résonance magnétique
J Radiol 2007;88:1689-94
Objectifs. La qualité des examens de cholangiopancréatographie par
résonance magnétique (CPRM) est fréquemment dégradée par le signal
élevé du contenu digestif sur les séquences fortement pondérées en T2.
L’objectif de cette étude est d’évaluer les résultats obtenus par l’ingestion
de jus d’ananas en tant qu’agent de contraste négatif en CPRM.
Matériels et méthodes. Nous avons comparé les résultats de la CPRM de
50 patients ayant ingéré du jus d’ananas et de 50 patients ayant ingéré
un produit de contraste paramagnétique (ferumoxsil-Lumirem®). Les
observateurs ne connaissaient pas la nature du produit de contraste
ingéré. La qualité des examens a été analysée en terme de suppression
du signal au sein de l’estomac, du cadre duodénal et du grêle proximal
et en terme de visualisation du canal pancréatique, des voies biliaires
intrahépatiques et de la voie biliaire principale. In vitro, nous avons
analysé l’intensité du signal de différents jus d’ananas disponibles sur le
marché, en se servant de séquences pondérées en T1, de séquences
pondérées en T2 et de séquences de CPRM. L’intensité du signal a été
corrélée avec la concentration de manganèse mesurée par spectrométrie
d’émission atomique. Finalement, les observateurs ont comparé les
goûts respectifs du jus d’ananas et du ferumoxsil.
Résultats. Sur les séquences de CPRM, les résultats étaient équivalents
en terme de suppression de signal de l’estomac, du cadre duodénal et du
grêle proximal après ingestion de jus d’ananas ou de ferumoxsil. La
visualisation du canal de Wirsung, des voies biliaires intrahépatiques et
de la voie biliaire principale était équivalente après ingestion d’ananas
ou de ferumoxsil. L’intensité du signal des différents jus d’ananas sur
les séquences pondérées en T2 et sur les séquences de CPRM était bien
corrélée avec la concentration de manganèse mesurée pour chacun des
jus par spectrométrie. D’importantes variations de la concentration de
manganèse ont été observées entre les différents jus d’ananas testés. La
concentration de manganèse variait entre 3,65 et 27,24 mg/L. Les
observateurs ont considéré que le jus d’ananas avait un laquo;bon» ou
un laquo;très bon» goût alors que le ferumoxsil avait un
laquo;mauvais» ou laquo;très mauvais» goût.
Conclusion. L’ingestion de jus d’ananas permet d’obtenir une diminution
efficace du signal du tractus digestif en CPRM d’une façon comparable à
celle qui est obtenue avec les produits de contraste paramagnétiques.
Comme la concentration de manganèse est largement variable dans les
différents jus d’ananas disponibles dans le commerce, un jus d’ananas à
haute concentration de manganèse doit être sélectionné.
Abstract
Purpose. The quality of magnetic resonance cholangiopancreato-
graphy (MRCP) images is frequently degraded by high signal from
the gastrointestinal tract on heavily T2W images. The purpose of this
study is to evaluate pineapple juice (PJ) as an oral negative contrast
agent in MRCP.
Materials and Methods. Results from MRCP in 50 patients with PJ
and 50 patients with paramagnetic contrast (ferumoxsil-Lumirem®)
were compared. Reviewers were blinded to the type of contrast agent.
Exam quality was recorded with regards to signal suppression in the
stomach, duodenum and proximal small bowel and with regards to
pancreatic duct and biliary ducts visualization. In vitro, the signal
characteristics of several commercially available brands of PJ were
assessed using T1W, T2W and MRCP sequences. Signal intensity
was correlated with the manganese concentration measured using
atomic absorption spectrometry. Finally, the reviewers compared the
taste of PJ and ferumoxsil.
Results. On MRCP sequences, results were similar with regards to
signal suppression in the stomach, duodenum and proximal small
bowel with PJ and ferumoxsil. Visualization of the pancreatic duct,
intrahgepatic bile ducts and CBD was similar with PJ and
ferumoxsil. The signal intensity of commercially available brands of
PJ on T2W and MRCP sequences correlated well with the measured
manganese concentration on spectroscopy. Variations in manganese
concentration were observed, with values ranging from 3.65 to 27.24
mg/L. The reviewers noted that PJ tasted “good” or “very good” and
that ferumoxsil tasted “bad” or “very bad”.
Conclusion. Ingestion of PJ provides effective signal suppression in
the GI tract on MRCP, similar to paramagnetic contrast agents.
Because manganese concentration is highly variable in commercially
available PJ brands, a brand with high manganese concentration
should be selected.

Mots-clés : Cholangiopancréatographie par résonance magnétique. Key words: MRCP. Contrast agents. Biliary tract, MR. Pancreatic
Agents de contraste. IRM, voies biliaires. IRM, canaux pancréatiques. duct, MR. Manganese. Atomic absorption spectrometry.
Manganèse. Spectrométrie d’émission atomique.

To cite the present paper, use exclusively the following reference. Arrivé L, Coudray C, Azizi L, Lewin M, Hoeffel C, Monnier-Cholley L,
Lacombe C, Vautier S, Poupon JL, Tubiana JM. Utilisation du jus d’ananas en tant qu’agent de contraste négatif en cholangiopancréatographie
par résonance magnétique (full text in english on www.masson.fr/revues/jr). J Radio 2007;88:1689-94.

(1) Service de radiologie, Assistance-Publique-Hôpitaux de Paris, Faculté de Médecine Pierre et Marie Curie, Hôpital Saint-Antoine, 184, rue du Faubourg Saint-Antoine, 75012 Paris.
(2) Laboratoire de toxicologie biologique, Groupe Hospitalier Lariboisière — Fernand Widal, 2, rue Ambroise Paré, 75010 Paris.
Correspondence : L Arrivé
E-mail : lionel.arrive@sat.ap-hop-paris.fr
* Conflits d’intérêt potentiels : aucun des auteurs de cet article n’a de lien avec l’industrie agroalimentaire, l’agriculture traditionnelle ou biologique.

© 2014 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 11/08/2014
 1690 Pineapple juice as a negative oral contrast agent L Arrivé and al.
in magnetic resonance cholangiopancreatography

he role of magnetic resonance cho-
T langiopancreatography (MRCP)
in the evaluation of the biliary
tract and pancreatic ducts is now well es-
tablished (1-5). The MR techniques used
to image the biliary tract and pancreatic
ducts are based on heavily T2W sequen-
ces to enhance signal from structures
containing stationary fluid and provide
anatomical representations of these
structures. Signal from fluid in the sto-
mach, duodenum and proximal small
bowel may interfere with visualization
of the biliary tract and pancreatic ducts.
As such, MRCP is routinely performed
in fasting patients and negative oral
contrast agents are sometimes used.
Most groups use paramagnetic oral
contrast agents that reduce signal from
gastric and bowel fluid by reducing the
T2 relaxation time (6, 7).
Blueberry juice, characterized by a high
manganese concentration, has also been
used to reduce signal from gastric and
bowel fluid (8-12). However, large
quantities of blueberry juice are not rea-
dily available in routine clinical practice.
In addition, ingestion of a large volume
of blueberry juice often is difficult be-
cause of its bitterness. Pineapple juice
(PJ), also characterized by a high manga-
nese concentration, has been proposed as
a negative oral contrast agent for MRCP
(13-15).
The purpose of this article was to compa-
re the efficacy of PJ as a negative oral
contrast agent for MRCP to that of a pa-
ramagnetic contrast agent, compare the
in vitro efficacy of different types of
commercially available PJ, and compare
the taste of these different contrast agents.
echo train to restore the residual transver-
se magnetization backup into the longitu-
dinal direction. The matrix was 256x256,
the slice thickness was 1 mm and the
voxel size was 1x1x1 mm. Images were
processed using a maximum intensity
projection algorithm (MIP) to generate
multiplanar reformatted anatomical re-
presentations. Free breathing acquisi-
tions usually range from 3 to 6 minutes
using the prospective acquisition and cor-
rection (PACE) feature. Patients were
instructed to drink 400 ml of PJ (total
manganese concentration: 18.73 mg/L) or
400 ml of ferumoxsil (Lumirem®) 15 mi-
nutes prior to scanning.
In vivo evaluation
Three reviewers blinded to the type of
oral contrast (ferumoxsil or PJ) evaluated
the quality of signal suppression of gastric,
duodenal and bowel fluid using a 5-point
scale, from “very poor” to “very good”
(table I). The observers then evaluated
the quality of biliary tract and pancreatic
duct visualization, also using a 5-point
scale, from “very poor” to “very good”
(table II).
Table I
Quality of signal suppression (results shown in %).
Signal Stomach
suppression
Lumirem® PJ
Very poor 0 0
Poor 6 6
Fair 4 4
In vitro evaluation
Fifteen tubes containing samples from
commercially available PJ (n=9), orange
juice (n=2), grape juice (n=1), blueberry
juice (n=1), water (n=1) and ferumoxsil
(Lumirem®) (n=1) were imaged using
the magnet previously described using
gradient-echo T1W (TR: 174, TE: 4),
HASTE T2W (TR: 1200, TE: 114) and
RARE MRCP (TR: 4160, TE: 1100) se-
quences. The samples were then analyzed
to determine the concentration of man-
ganese. The samples were mineralized
in the presence of heated concentrated
nitric acid by microwave-assisted
digestion.
Total manganese dose measurements
were performed by atomic absorption
spectrometry using induction plasma.
The total manganese concentration for
each sample was compared to the corres-
ponding MR signal intensity.
Taste evaluation
Ten observers then evaluated the taste of
PJ and ferumoxsil. The observers were
blinded to the nature of the proposed pro-
duct for review and taste tests for both
Duodenum Proximal bowel
Lumirem® PJ Lumirem® PJ
0 0 2 0
0 4 2 6
6 8 16 4

Good 30 20 14 12 12 20
Materials and Methods Very good 60 70 80 76 68 70

Magnetic resonance imaging

MRCP examinations were performed
on a 1.5T MR unit (Magnetom Sympho-
Table II
ny, Siemens, Erlangen, Germany) using Quality of visualization (results shown in %).
the body coil and a six-element phased
Visualization Pancreatic Intrahepatic Suprapancreatic Intrapancreatic
array coil. The imaging protocol inclu- duct bile ducts CBD CBD
ded an axial breath-hold gradient-echo
Lumirem® PJ Lumirem® PJ Lumirem® PJ Lumirem® PJ
T1W (TR/TE: 174/4) sequence, an axial
non-breath-hold non-fat-suppressed Very poor 2 0 2 0 0 0 0 0
FSE T2W (HASTE – TR/TE msec:
Poor 18 12 6 0 4 0 2 0
1200/114; matrix 176x256; slice thic-
kness 6 mm) sequence and a high-resolu- Fair 18 6 18 8 4 4 2 4
tion 3D FSE sequence with very long TE Good 16 20 20 14 22 10 26 16
(TR/TE: 1400/800). A restoration pulse
less than 90 was applied at the end of the Very goode 46 62 54 78 70 86 70 80

J Radiol 2007;88

© 2014 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 11/08/2014
 1691 Pineapple juice as a negative oral contrast agent L Arrivé and al.
in magnetic resonance cholangiopancreatography

products were not performed on the same
dates. A 5-point scale was used: very bad,
bad, drinkable, good, and very good.
Statistical analysis
All statistical comparisons either for qua-
lity of signal suppression, quality of bilia-
ry tract and pancreatic duct visualization
or taste of different oral contrast prepara-
tions were performed with the chi-2 test
for non-paired series using a p<0.05 for
statistical significance.
Results
Signal suppression in the stomach, duo-
denum and proximal small bowel was
“good” or “very good” in most cases with
either PJ or ferumoxsil (table I). There
was no significant difference regarding
the efficacy of signal suppression
between PJ and ferumoxsil. Visualiza-
tion of the main pancreatic duct, intra-
hepatic bile ducts, suprapancreatic CBD
and intrapancreatic CBD was conside-
red “good” or “very good” in most cases
(table II). There was no significant diffe-
rence regarding the quality of visualiza-
tion of the biliary tract or pancreatic
duct.
The observers noted that signal from GI
tract structures did not interfere with the
visualization of the main pancreatic duct,
intrahepatic bile ducts, suprapancreatic
CBD and intrapancreatic CBD irrespec-
tive of PJ or ferumoxsil ingestion (fig. 1).
The difference between both agents was
not significant.
On gradient-echo T1W sequences, si-
gnal from the stomach after ferumoxsil
ingestion was usually hypointense relative
to spleen (fig. 2a) whereas signal from
the stomach after PJ ingestion was
always hyperintense relative to liver
(fig. 3a).
On HASTE T2W sequences, signal from
the stomach after ferumoxsil ingestion
was usually hypointense relative to mus-
cle (fig. 2b) whereas signal from the sto-
mach after PJ ingestion was usually
between that of liver and spleen or spleen
and adipose tissue (fig. 3b).
Total manganese concentration was es-
sentially zero for water and less than
1 mg/L for orange juice, grape juice and
ferumoxsil. Total manganese concen-
tration in blueberry juice was 41 mg/L.
Fig. 1:
Total manganese concentration for PJ
was variable between brands (minimum
3.6 mg/L – maximum 27.2 mg/L)
(table III).
MR imaging of the different samples
showed that the efficacy of signal sup-
pression from GI structures on MRCP
images was sufficient above a concentra-
tion of 15 mg/L (fig. 4). Signal suppres-
sion was insufficient for PJ brands with
manganese concentrations of 3.6, 6.8 and
12 mg/L respectively.
The seven brands of PJ with manganese
concentration above 10 mg/L were all
made from concentrate and from diffe-
rent manufacturers. PJ juice in vivo had a
manganese concentration of 18.7 mg/L.
Both PJ brands with low total manganese
concentrations (3.6 and 6.8 mg/L) were
“pure squeezed” juices from standard or
organic farming.
3D MRCP image (TR/TE 1400/
800) after ingestion of PJ.

a b
Fig. 2: Axial gradient-echo T1W image (TR/TE : 174/4) after ingestion of ferumoxsil (Lumirem®) .
a Gastric signal is hypointense relative to muscle. On HASTE T2W image,
b Gastric signal is hypointense relative to muscle.

J Radiol 2007;88

© 2014 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 11/08/2014
 1692 Pineapple juice as a negative oral contrast agent L Arrivé and al.
in magnetic resonance cholangiopancreatography

a b
Fig. 3: Axial image after ingestion of PJ. On gradient-echo T1W image (TR/TE : 174/4).
a Gastric signal is hyperintense relative to liver. On HASTE T2W image (TR/TE : 1200/114).
b gastric signal is hypointense relative to spleen.

Table III
Total manganese concentration from different samples.
Sample Nature Mn total (mg/L)

1 PJ
2 PJ
3 PJ
4 PJ
5 PJ
6 PJ
7 PJ
8 PJ
9 PJ
10 Orange J.
11 Grape J.
12 Blueberry J.
13 Orange J.
14 Water
15 Ferumoxsil
Brand of different samples:1. Leader Price – 2. Réa – 3. Mini fruité – 4. Pressade – 5. Monoprix pur jus – 6. Vitalia –
7. Carrefour – 8. Auchan – 9. Monoprix.
Taste for PJ was considered “very good”
or “good” in 80% of cases. Taste for feru-
moxsil was considered “bad” or “very
bad” in 80% of cases (p<0.05).
Discussion
MRCP has established itself as a prefer-
red modality for morphological evalua-
tion of the biliary tract and pancreatic
duct (1-4).
MRCP is based on heavily T2W sequen-
ces to enhance signal from stationary
11.81
18.73
15.24
27.24
3.65
6.83
17.28
18.84
21.72
0.38
0.73
40.68
0.40 Fig. 4: MRCP image (TR : 4160, TE :
0.03 1100) from different samples. Sig-
nal from the 6 PJ brands with
0.56 manganese concentration above
15 mg/L is suppressed. Signal
from the blueberry juice and feru-
moxsil is suppressed. Signal is
still present for PJ brands with
manganese concentrations of
12 mg/L (1), 3.6 mg/L (2), and
fluids. Signal from fluid in the stomach, 6.3 mg/L (3) and the orange juice
duodenum and proximal small bowel (4, 6), grape juice (5) and water (7)
samples.
may interfere with visualization of the bi-
liary tract and pancreatic ducts. As a re-
sult, several investigators have looked at
negative oral contrast agents to suppress
signal from GI structures (6, 7).
The agent currently most frequently used rized by a high manganese concentration
is the paramagnetic contrast agent feru- (8-12).
moxsil (Lumirem®) consisting of a sus- We have demonstrated in this study that
pension of superparamagnetic iron oxide PJ was an attractive alternative with good
causing marked shortening in T2 relaxa- suppression of signal from GI tract struc-
tion time (6, 7). Other agents have been tures. Visualization of the main pancrea-
used, including blueberry juice characte- tic duct, intrahepatic bile ducts, and CBD

J Radiol 2007;88

© 2014 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 11/08/2014
 L Arrivé and al.
was considered of “good” or “very good”
quality in most cases with very little inter-
ference from bowel structures. Results
with PJ and ferumoxsil were similar.
Suppression of signal from GI structures
was effective only on heavily T2W MR-
CP sequences. Signal suppression was not
as effective on less heavily T2W sequen-
ces such as HASTE sequences.
We have also demonstrated in vitro that
manganese concentration from different
brands of commercially available PJ was
variable, with two brands of “pure squee-
zed” PJ juice containing a manganese
concentration too low to result in clinical-
ly useful signal suppression.
Finally, we have demonstrated that the
taste of PJ was significantly better tolera-
ted than the taste of ferumoxsil.
We are not aware of any complication re-
lated to the ingestion of a reasonable volu-
me of PJ except for the associated sugar
load whereas adverse reactions may occur
with ferumoxsil, including oral paresthe-
sia due to the concentration of high iron
oxide, and gastrointestinal symptoms:
diarrhea, abdominal cramps, bloating
and flatulence.
In addition, even if the price of a bottle of
Lumirem® is reasonable (15.31 euros), it
remains markedly more expensive than a
bottle of PJ billed at 0.20 euro in our insti-
tution’s kitchen.
Blueberry juice has been used as a negati-
ve oral contrast agent for MRCP (8-12).
We have confirmed the high concentra-
tion of manganese in blueberry juice,
higher than for PJ. However, this higher
concentration is not an advantage by itself
since a similar clinical effect can be achie-
ved by using an agent with lower manga-
nese concentration. It should be noted
that given its very high manganese
concentration, blueberry juice could be
diluted to 1/4, which could make it more
clinically usable and would also presuma-
bly reduce its bitter taste.
Some authors have proposed a mixture of
PJ and Gadolinium to suppress GI signal
(15). Reported results were good, but the
mixing procedure is more complicated.
Other authors have proposed the use of
dilute Gadolinium as an oral contrast
agent to reduce GI signal. Preliminary
data from a patient population of 23
showed that this technique appeared ef-
fective and well tolerated (16).
Our study presents several limitations.
The administration of ferumoxsil and PJ
was not randomized. We did not verify if
J Radiol 2007;88
© 2014 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 11/08/2014
Pineapple juice as a negative oral contrast agent
in magnetic resonance cholangiopancreatography
all patients were fasting or had ingested
the entire volume of negative contrast
agent. This probably explains the rare
cases of poor results, especially with feru-
moxsil suggesting that some patients may
not have ingested the entire volume of
contrast.
Even though the reviewers were blinded
to the type of negative oral contrast agent
in any individual subject, they could sus-
pect which agent was used by reviewing
the signal intensity of gastric fluid on
T1W images.
Similarly, even though reviewers during
the tasting phase were blinded to the na-
ture of the ingested agent, they could
easily deduct which contrast was under
review based on evaluation of color and
taste.
Finally, because of their high manganese
concentration, PJ and blueberry juice are
characterized by T1 relaxation time shor-
tening causing fluids to appear hyperin-
tense on T1W sequences. As a result, this
type of contrast agent would be of limited
value on MRCP acquisitions obtained
following the intravenous administration
of a contrast agent with short T1 and bi-
liary excretion where T1W imaging
would be required. The use of PJ or blue-
berry juice would increase signal from GI
structures and interfere with the inter-
pretation (17).
Conclusion
We believe that PJ is an attractive negati-
ve contrast agent for MRCP as long as the
selected brand has a manganese concen-
tration that is sufficient.
References
1. Yu J, Turner MA, Fulcher AS, Halvor-
sen RA. Congenital anomalies and nor-
mal variants of the pancreaticobiliary
tract and the pancreas in adults: part 1, bi-
liary tract. AJR Am J Roentgenol 2006;
187:1536-43.
2. Yu J, Turner MA, Fulcher AS, Halvor-
sen RA. Congenital anomalies and nor-
mal variants of the pancreaticobiliary
tract and the pancreas in adults: part 2,
pancreatic duct and pancreas. AJR Am J
Roentgenol 2006;187:1544-53.
3. Hoeffel C, Azizi L, Lewin M, Laurent V,
Aube C, Arrivé L, et al. Normal and pa-
thologic features of the postoperative bi-
1693
liary tract at 3D MR cholangiopancreato-
graphy and MR imaging. Radiographics
2006;26:1603-20.
4. Pilleul F, Courbière M, Henry L, Scoa-
zec JY, Valette PJ. La cholangio-IRM
dans le diagnostic étiologique des sténo-
ses biliaires : corrélation anatomopatho-
logique. J Radiol 2004;85:25-30.
5. Regent D, Laurent V, Meyer-Bisch L,
Barbary-Lefèvre C, Corby-Ciprian S,
Mathias J. La douleur biliaire comment
la reconnaître ? comment l’explorer ? J
Radiol 2006;87:413-29.
6. Kim YK, Kim CS, Lee JM, Ko SW,
Chung GH, Lee SO, et al. Value of ad-
ding T1-weighted image to MR cholan-
giopancreatography for detecting intra-
hepatic biliary stones. AJR Am J
Roentgenol 2006;187:W267-74.
7. Broglia L, Tortora A, Maccioni F, Arpe-
sani R, Marcelli G, Ascarelli A, et al. Op-
timization of dosage and exam technique
in the use of oral contrast media in ma-
gnetic resonance. Radiol Med 1999;97:
365-70.
8. Papanikolaou N, Karantanas A, Maris T,
Gourtsoyiannis N. MR cholangiopancrea-
tography before and after oral blueberry
juice administration. J Comput Assist
Tomogr 2000;24:229-34.
9. Hiraishi K, Narabayashi I, Fujita O, Ya-
mamoto K, Sagami A, Hisada Y, et al.
Blueberry juice: preliminary evaluation
as an oral contrast agent in gastrointesti-
nal MR imaging. Radiology 1995;194:
119-23.
10. Karantanas AH, Papanikolaou N, Kalef-
Ezra J, Challa A, Gourtsoyiannis N.
Blueberry juice used per os in upper ab-
dominal MR imaging: composition and
initial clinical data. Eur Radiol 2000;
10:909-13.
11. Asbach P, Breitwieser C, Diederichs G,
Eisele S, Kivelitz D, Taupitz M, et al. Ci-
ne magnetic resonance imaging of the
small bowel: comparison of different oral
contrast media. Acta Radiol 2006;47:899-
996.
12. Rivoal E, Letourmy JM. Using blueberry
juice as a negative contrast agent in pae-
diatric MR urography. Pediatr Radiol
2000;30:498.
13. Riordan RD, Khonsari M, Jeffries J, Mas-
kell GF, Cook PG. Pineapple juice as a
negative oral contrast agent in magnetic
resonance cholangiopancreatography: a
preliminary evaluation. Br J Radiol 2004;
77:991-9.
14. Schreyer AG, Herfarth H, Kikinis R,
Seitz J, Schölmerich J, Geissler A, et al.
3D modeling and virtual endoscopy of
the small bowel based on magnetic reso-
nance imaging in patients with inflam-
matory bowel disease. Invest Radiol
2002;37:528-33.
 1694 Pineapple juice as a negative oral contrast agent L Arrivé and al.
in magnetic resonance cholangiopancreatography

15. Coppens E, Metens T, Winant C, Matos C.
Pineapple juice labeled with gadolinium: a
convenient oral contrast for magnetic reso-
nance cholangiopancreatography. Eur Ra-
diol 2005;15:2122-9.
16. Chan JH, Tsui EYK, Yuen MK, Szeto ML,
Luk SH, Wong KP, et al. Gadopentetate
dimeglumine as an oral negative gastroin-
testinal contrast agent for MRCP. Abdom
Imaging 2000;25:405-8.
17. Assaban M, Aube C, Lebigot J, Ridereau-
Zins C, Hamy A, Caron C. Intérêt de
l’IRM avec perfusion de mangafodipir
trisodium dans la détection des fuites bi-
liaires. J Radiol 2006;87:41-7.

J Radiol 2007;88

© 2014 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 11/08/2014