Image Evaluation in Magnetic Resonance Cholangiopancreatography
K. E. P. Pinho
1a
, P. M. Gewehr
1b
, A. C. Pinho
2c
, A. M. Gusso
3d
and C. A. Goedert
3e
1
Graduate Program of Electrical and Computer Engineering (CPGEI)
from Federal University of Technology-Parana (UTFPR), Curitiba, Brazil
2
Departament of Electrical Engineering (DAELT), UTFPR, Curitiba, Brazil
3
Diagnostic Clinic (Cetac), Curitiba, Brazil
Keywords: Magnetic Resonance Cholangiopancreatography, Contrast Agent, Image Evaluation, Image J, Natural Juice.
Abstract: The objective of this study is to evaluate the image quality with two different contrast agents in MRCP
compared to medical evaluation and by using the software Image J®. Natural juices and pulps of different
types (açai liquid and powder; and blend) were selected. The selection of patients (31 women and 33 men)
was performed at Clinical Hospital, which provides general care in Curitiba city (Brazil). The application of
the MRCP protocol followed a sequence tested in healthy volunteers and for the samples described. For image
analysis, 2 radiologists participated and were identified as evaluator 1 (E1) and 2 (E2), in order to identify the
effect of the contrasts on the images. For the 6 samples tested, only 2 samples remained dark on T2 weighting,
which prevents their use as contrast agent. The evaluation of the images was performed separately for each
evaluator on different days and places, to identify an appropriate action for the contrasts (A and B). The use
of the software (Image J®) allowed a less subjective analysis of the image quality when compared to the
evaluation of radiologists and, for the examples presented, a quantitative assessment since the chosen images
were submitted to the software analysis.
1 INTRODUCTION
The acquisition of images by magnetic resonance
(MRI) has the advantage to turn easier the views of
soft tissues as well as organs of difficult access
compared to other imaging methods (e. g., X-rays and
computerized tomography). However, to obtain
quality images, it is important that they can show
areas of intense (white), weak (dark) and intermediate
(gray levels) signals (Westbrook, Roth and Talbot,
2011;
Jornada, Murata and Medeiros, 2016).
Among the types of MR image acquisitions, this
study is concerned to Magnetic Resonance
Cholangiopancreatography (MRCP). This technique
makes use of a negative contrast to identify and to
show parts of the gastrointestinal system, in particular
the images of the pancreas and gall- bladder which
are superimposed (Xiao and Zhang, 2010).
The
contrast to be used can be manufactured or
a
https://orcid.org/0000-0002-3542-9283
b
https://orcid.org/0000-0002-9694-7906
c
https://orcid.org/0000-0002-4014-8072
d
https://orcid.org/0000-0003-0716-1002
e
https://orcid.org/0000-0002-5255-4844
natural.
The use of a natural contrast, as the juice of
some fruit, is more indicated since it does not cause
adverse reactions and present an image similar to a
manufactured one (Fraga et al., 2004; Pinho et al.,
2019).
A
juice as contrast agent must be
paramagnetic;
act as biphasic contrast (showing up
positive for T1 sequences and, negative for T2);
be
uniformly distributed in the digestive cavity and small
intestine;
non-toxic;
and be
affordable (Duarte,
Furtado and Marroni, 2012).
The contrast agents
should have some metals that help in identifying the
image, such as iron (Fe) and manganese (Mn).
These
can be found in juices (pineapple, açai and blueberry),
teas and specific preparations (Ghanaati et al., 2011;
Griffin, Edwards and Grant, 2012; Renzulli et al.,
2019).
Oral contrast agents in MRCP examinations must
present a low signal in weighted T2, with negative
98
Pinho, K., Gewehr, P., Pinho, A., Gusso, A. and Goedert, C.
Image Evaluation in Magnetic Resonance Cholangiopancreatography.
DOI: 10.5220/0008987200980105
In Proceedings of the 13th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2020) - Volume 4: BIOSIGNALS, pages 98-105
ISBN: 978-989-758-398-8; ISSN: 2184-4305
Copyright
c
2022 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
contrast effect in the region impregnated with the
contrast (Mantau et al., 2014).
In the case of the
MRCP, it is important to eliminate the stomach
(Figure 1, arrow 1) duodenum signals (Figure 1,
arrows 2) and to facilitate the visualization of gall-
bladder (Figure 1, arrow 3), common bile (Figure 1,
arrow 4a) and pancreatic ducts region (Figure 1,
arrow 4b).
The objective of this study is to evaluate the image
quality with two different contrast agents in MRCP
compared to medical evaluation and its confirmation
by the software Image (Souza et al., 2014). This
software was used to verify if both contrasts could
present overall MRCP equivalent image quality.
Figure 1: MRCP image after negative oral contrast
administration. Arrows indicate: (1) the stomach region and
(2) the duodenum, both erased by contrast action, (3) the
gallbladder and (4a) the common bile duct and (4b)
pancreatic duct region.
2 MATERIALS AND METHODS
The research project was approved by UTFPR Ethics
Committee (number 02.520.512.0.0000.5547.
The selection of patients (31 women and 33 men)
was performed at the Clinical Hospital of the Federal
University of Paraná (UFPR), which provides general
care in Curitiba city (Brazil).
The patients were
selected from
the
outpatient clinics of liver and fat
correlated diseases of the Hospital. Experiments were
produced to select the juice and MRCP exams were
performed in a Diagnostic Clinic of Curitiba city.
2.1 MRCP Protocol
2.1.1 Experiments with Phantom
Initially, natural juices and pulps of different types
(açai
liquid and powder; and blend) were selected.
Afterwards, they were placed in a water container, as
shown in Figure 2, and tested on a 1.5T MRI system
from General Electric Company (GE), model HDXT
with 12 channels, GE Healthcare Advantage
workstation running Centricity DICOM Viewer
version 3.0 software in the Clinic above. The protocol
determination of the samples was the same as that
used by the clinic for the examination of MRCP.
Initially: localizer (LOC) in 3 orthogonal planes (PL)
following Single-Shot (SS), Fast Spin Echo (FSE) in
apnea (LOC 3 PL SSFSE Apnea); radial cholangio
and axial lava T1 without fat (Pinho et al., 2014;
Pinho et al., 2018).
The parameters of the T1 weighting protocol
were: LAVA TR/TE 4.2/2.0 ms; FOV: 36x32 mm;
320x160; slice thickness: 4.0 mm and 0.70 NEX;
inversion time (TI): 7.0 ms. For T2: acquisition in
Fast Imaging Employing Steady-State
(Fiesta)/TR/TE 4.4/2.0 ms; FOV: 36x36 mm;
224x320; slice thickness: 3.0 mm; 1.0 NEX:
inversion time: 200 ms (Westbrook, 2010; Pinho et
al., 2018).
Figure 2: T2 sequence of image of açai samples and
commercial contrast. Identification was performed from
left to right, with numbers 1, 2 and 3 above, and 4, 5 and 6
below.
The phantom tests were carried out with six
samples of açai juice of different brands and with
different water concentrations. Figure 2 illustrates
these samples, which can be identified as follows,
respectively:
1
2
4 a
3
4 b
1 2
5
4
3
6
Image Evaluation in Magnetic Resonance Cholangiopancreatography
99
1- Açai 100 g /100 mL of water;
2- Açai powder with 13 g /100 mL of water;
3- Commercial contrast;
4- Açai frozen with 100 g /100 mL of water;
5- Açai more concentrated, 100 g/80 mL of water;
6- Açai powder with 6g /100 mL of water.
2.1.2 Experiments with Patients
The MRCP examination protocol was the same as for
the samples of juices described above, besides the
acquisition of multiple thin slices in the coronal plane
for this work: Half Acquisition Single Shot Turbo
Echo (HASTE), Turbo Spin Echo (TSE), following
thick radial cuts in FSE/TSE also with strong
weighting in T2. Here the cutting plan is directed to
the distal common bile duct.
The acquisitions were
made in Axial 2D FIESTA (with fat saturation) Array
Spatial Sensitivity Encoding Technique (ASSET)
and the radial cholangio sequence for two days of
tests, in order to compare the effectiveness of
contrasts (Sanchez et al., 2009; Pinho et al., 2018).
Each patient was identified by the respective
gender after a cardinal number (female, male), e.g.:
F1, M2 to facilitate image acquisition and storage.
The application of the MRCP protocol followed a
sequence tested in healthy volunteers. In addition,
doctors were available to perform MRCP exams in
that clinic. Since the patients would need the
examination report, on the first day a commercial
contrast (labeled A) was administered with a total
abdominal sequence and administration. On the
second day the natural contrast of açai juice (labeled
as B) was administered, and after the sequence of
MRCP was started. The dose of each contrast was 200
mL divided into 2 portions of 100 mL, one dose was
given after the anamnesis and another 10 minutes
later (Pinho et al., 2018).
2.2 Image Evaluation
For image analysis, 2 radiologists participated and
were identified (this study) as evaluator 1 (E1) and 2
(E2), both with experience in the field of diagnostic
imaging, in order to identify the effect of the contrasts
on the images.
After the exams were completed, the images were
saved in the PACS (Picture Archiving and
Communication System) (Marques et al., 2005)
system and available on file so that each evaluator
could access and analyze the effects of the contrasts
(commercial and natural), according to the scale from
1 to 4. By observing the regions of interest for
examining MRCP, i.e,. the oral contrast should erase
the sign of the stomach and duodenum. Score 1 means
that there is a hyperintensity signal of stomach and
duodenum and it is not possible to evaluate these
structures.
Score 2: assessment occurs when there is
a partial view of the structures.
In score 3,
hyperintensity signal does not hinder the analysis of
the structures, and score 4 means that there is no
signal hyperintensity for stomach and duodenum,
which makes clearer the MRCP image (Duarte,
Furtado and Marroni, 2012; Pinho et al., 2019).
Image software
(obtained free from
http://imagej.nih.gov
) was employed to analyze and
compare the image quality of the patients by
separating a common bile duct region, with
the same
dimension (selection rectangle with size
approximately 97.85 mm x 2.58 mm (length and
height)), in Figures 4, 5, 6 e 7. The chosen sections of
the Figures were selected by the radiologists and used
later for the Image
estimation of gray levels since
those regions are used for medical evaluation in order
to detect physiological alteration or correlated
diseases. For the construction of the figures with
Image
, the length corresponds to the anatomical
region of interest and the value to gray levels
(intensity pixels).
For that, it was chosen among the
assessed medical images, which ones
showed
a coincidence between the two MRCP
sequences for the two contrasts, and about same
scores provided by the evaluators (Brianezi, Camargo
and Miot, 2009; Pinho et al., 2019).
3 RESULTS
3.1 Experiments with Phantom
For the 6 samples tested, only 2 (sample 2 of açai
powder and 5 of açai more concentrated) remained
dark on T2 weighting, which prevents their use as
contrast agent. It is observed that in the T2 sequence
presented, the dark images of the juices and mixtures
provided the details required in the MRCP
examination, since for the bile and pancreatic images
acquisition, the administered contrast must eliminate
the residual signal.
The images of açai samples taken from Figure 2
were handled and their Regions of Interest (ROI)
values were measured. The best samples were 1, 3, 4
and 6 for T1 and T2 weighting, as seen in Figure 3.
ROI values of T1 weighting should be as great as
possible, and the values obtained were 1038
(samples 1, 3, 4 and 6). The value of 527.6 of
substance 3 is not compatible in T2 since it was low
BIOSIGNALS 2020 - 13th International Conference on Bio-inspired Systems and Signal Processing
100
compared to the others. Also in Figure 2 for T2, ROI
values of the samples must be higher because the
expected behavior is present in the image (darker),
and the amounts obtained were 1037.7 (sample 1);
527.6 (2); 1111.8 (3); 1435.1 (4); 419.4 (5) and 1188.
2 (sample 6).
The best samples for T2 were 1, 3, 4 and 6. It was
decided to use the sample 1 for ease of acquisition and
availability.
3.2 Patients and Image Analysis
The application of the MRCP protocol for 64 patients
produced 12 acquisitions were obtained with contrast
A as well as 12 with contrast B (1x12x12=144 MRCP
images). In order to present the analysis method using
Image J® to compare the images and, considering the
individual characteristics of patients, quantity and
complexity of information provided by the number of
images for each patient on 2 days of exams, it was
decided to present the results for 2 patients (M43 and
M5). One case is considered more complex and the
other is simple, but both need the action of the
contrasts to show the organs and/or to observe
existent diseases when pertinent and thus proving that
Image J® can be used for all practical cases.
Image Evaluation
The evaluation of the images was performed
separately for each evaluator on different days and
places; they have chosen within the MRCP sequence
two images to identify an appropriate action for the
contrasts (A and B), one for each type of contrast.
Figure 3 shows an image acquired with contrast A
and, Figure 4 an image acquired with contrast B for
the radial cholangio sequence of patient M43. Both
images had scores 3 from evaluators E1 and E2 on 2
days. The medical report for patient M43 described a
small ascites in the hepatic region (more visible in
Figure 3, arrow 1), chronic liver disease, distended
gallbladder and with thickened walls, with better
anatomic view in Figure 4 (arrow 2).
It is noted that on the two images (Figure 3 and 4)
there was reduced signals from the stomach and
duodenum, showing the complete bile duct. The
Figures above have shown adequate white levels
(gallbladder region, arrows 2 and, common bile duct,
arrows 3) and dark levels for stomach and duodenum
(Figure 1, arrows 1 and 2), as must be present for a
quality medical report of MRCP.
Figure 5 shows an image acquired with contrast A
and, Figure 6 an image acquired with contrast B for
the radial cholangio sequence of patient M5. Both
images had scores 3 from evaluators E1 and E2 on
first day for contrast A. On second day, considering
contrast B, E1 assigned score 3 and, E2 score 4. The
medical report described that fat liver and the other
organs (gallbladder, pancreas and ducts) were with
normal anatomic aspect.
Figure 3: Image from patient M43 which had a score 3 with
contrast A. Arrows indicate: (1) ascites region; (2)
gallbladder; (3) common bile duct. The rectangle indicates
the area of the common bile duct, chosen for the software
Image J®.
The Figures 5 and 6 have shown both contrast
with pretty similar images for MRCP, exception for
organs without interest as kidney and large intestine.
The images (Figures 5 and 6) show adequate white
levels (gallbladder region, arrows 2 and, common bile
duct, arrows 3). Particularly for Figure 5 (arrow 1),
the stomach region presented some points with white
levels where it should be dark.
The software Image J® was applied to Figures 3, 4, 5
and 6 generating a quantitative analysis of a selected
section of the common bile duct, close to the
duodenum. The arrows indicate the selected areas of
approximately 97.85 mm x 2.58 mm. The areas were
taken to build the curves of Figures 7 and 8 which
show the gray levels (pixels) as a function of distance
(width duct) for contrasts A and B. For both Figures
(7 and 8), arrows (1) indicate the duodenum and
gallbladder region, arrows (2) the common bile duct,
arrows (3) the pancreatic duct and, arrows (4) the
pancreas region.
Figure 7 shows curves as obtained with data from
Figures 3 and 4 and the software Image J®. For
regions (1) e (2) a peak value with gray levels (600.8)
1
2
3
1
1
Image Evaluation in Magnetic Resonance Cholangiopancreatography
101
Figure 4: Image from patient M43 which had a score 3 with
contrast B. Arrows indicate: (1) ascites region; (2)
gallbladder; (3) common bile duct. The rectangle indicates
the area of the common bile duct, chosen for the software
Image J.
Figure 5: Image from patient M5 which had a score 3 with
contrast A. Arrows indicate: (1) duodenum region; (2)
gallbladder; (3) common bile duct. The rectangle indicates
the area of the common bile duct, chosen for the software
Image J®.
was obtained at position 11.6, due to the presence of
ascites around the gallbladder (see Figure 3,
rectangle) as obtained with contrast A. Considering
the same region for contrast B, the values were quite
different. For region (3) the peak values were 365.5
(contrast A) and 284.5 (contrast B) close to point
50.3. For points from 0 to 46.4 (regions 1 and 2), a
correlation coefficient of 0.13 was calculated for the
Figure 6: Image from patient M5 which had a score 3 with
contrast B (E1) and score 4 (E2). Arrows indicate: (1)
duodenum region; (2) gallbladder; (3) common bile duct.
The rectangle indicates the area of the common bile duct,
chosen for the software Image J®.
gray levels of both contrasts. This is related to ascites
which caused some blanching over the image (Figure
4 arrow 1).
The average intensities of the same points for
contrast A was 316.6 and 113 for contrast B.
Considering region (3) which is important for an
MRCP medical report, the correlation between gray
levels was 0.95 among points from 46.4 to 54.1. The
average values for this region were 216.1 for contrast
A and, 166.4 for contrast B.
Figure 8 shows the curves as obtained with data
from Figures 5 and 6 and the software Image J®. It
shows (Figure 8) great similarity between curves for
contrasts A and B. The correlation coefficient among
all points (0 to 100) for gray levels was 0.92. The
average gray levels for points (0 to 46.4) were 32.2
and 35.7 for contrasts A and B, respectively. Taking
regions (2) and (3) from points 47.4 up to 60.6mm,
the average intensity for contrast A was 80.7 and 74.8
for contrast B.
4 DISCUSSION
The ROI values obtained for the açai samples showed
that for T2 with the higher values, they can be used as
contrast agent in MRCP, according to other authors
(Fraga et al., 2004; Espinosa et al., 2006 and Pinho et
al., 2019). Note that the açai value (sample 1) was
close to the commercial contrast (sample 3).
2
3
1
1
2
3
1
2
3
BIOSIGNALS 2020 - 13th International Conference on Bio-inspired Systems and Signal Processing
102
Two radiologists evaluated the images obtained in
MRCP exams giving notes from 1 to 4. Two MRCP
examples were selected (patient M43 e M5) to
produce a quantitative analysis employing Image J®.
For the example (M43) presented, the images
received the same notes (3) from both evaluators,
using the commercial contrast (A) and açai juice (B).
For case M5, both images had scores 3 from
evaluators E1 and E2 on 1
st
day for contrast A. On 2
nd
day, considering contrast B, E1 assigned score 3 and,
E2 score 4.
The chosen images, by both evaluators for M43
patient, are equal within the MRCP sequences.
Considering Figure 3, the gallbladder did not present
all the contours, differing from Figure 4 (where it
showed). This organ presents great motility with time
what can explain the differences between Figures 3
and 4 (arrows 2), besides the variability of food
ingestion, which alters gallbladder physiology.
For patient M43 (Figure 7), Image presented
gray levels with peak value of 600.8 and average
value of 316.6 for regions 1 and 2 on first day. These
values are bigger than those presented on second day.
The higher values are explained by means of Figure 3
since the ascites was more visible. The correlation
coefficient between contrasts curves was pretty low
for regions 1 and 2 due to the high variability of
ascites viewing. Taking only region 3, the correlation
is very high (0.95), what shows that both contrasts are
acting very well within the common bile duct.
Taking Figure 8 (Image J®) for M5 patient, there
were clean regions showing the common bile duct,
with gray average intensity of 32.2 (contrast A) for
region 1 (duodenum and gallbladder) and 35.7
(contrast B). For regions 2 and 3, which are important
to find out anatomical alteration or diseases, average
intensity was 80.7 (contrast A) and 74.8 (contrast B)
meaning that both contrasts had similar values and
clear views of the common bile and pancreatic ducts
regions, as expected for an oral negative contrast.
Also, evaluators noticed no difference between
contrasts.
The limitations of the work are related to the fact
that there is no known association between the
artifacts present in the image and the evaluators'
grades. Artifacts can be from the equipment (lack of
quality control) and/or from the patient.
The artifacts
from the patient may be due to illness, lack of
adequate preparation for the exam, physiological
changes and others. These modify the image quality,
as well as the note of the evaluator.
Image software showed for the cases
described, the one with presence of diseases there was
increase of average gray level (patient M43) for
ascites (about 113), while for patient M5 (normal
MRCP) the average was low (about 36).
5 CONCLUSIONS
The present study aimed to evaluate the image quality
with two different contrast agents (commercial and
Figure 7: Plots of the selected regions of radial cholangio images from the common bile duct of patient M43 as obtained with
the software Image J® The curves represent the gray levels against distance for contrasts A and B. Arrows indicate: (1)
duodenum region and gallbladder; (2) near common bile duct; (3) common bile duct and, (4) pancreas body.
0
100
200
300
400
500
600
0 102030405060708090100
Gray levels
Width distance (mm)
Contrast A
Contrast B
1
3
4
2
Image Evaluation in Magnetic Resonance Cholangiopancreatography
103
Figure 8: Plots of the selected regions of radial cholangio images from the common bile duct of patient M5 as obtained with
the software Image J®. The curves represent the gray levels against distance for contrasts A and B. Arrows indicate: (1)
duodenum region and gallbladder; (2) common bile duct; (3) pancreatic duct and, (4) pancreas body.
natural) in MRCP compared to medical evaluation
and confirmation by the software Image J®.
The natural contrast (açai) was able to erase the
signal from the stomach and duodenum (as shown),
as well as enhanced signal to the common bile duct,
as it should be in the clinic for a quality image in
MRCP.
The use of the software (Image J®) allowed a less
subjective analysis of the image quality when
compared to the evaluation of radiologists and, for the
examples presented (patients M43 and M5), a
quantitative assessment since the chosen images were
submitted to the software analysis. Thus, the gray
levels intensities obtained with the software Image
corroborates the view of the evaluators E1 and E2
with the images. Anyhow, the results open an
excellent opportunity for further studies, to establish
an automated protocol that uses available software
since it was useful to confirm the existence or not of
anatomical alteration and/or diseases.
ACKNOWLEDGEMENTS
To Araucaria Foundation from Parana State, for
providing research support through project number
355/2012.
To Clinics Hospital-UFPR and, Curitiba
Diagnostic Clinic, where this work was carried out.
REFERENCES
Brianezi, G., Camargo, J.L.V., Miot, H.A., 2009.
Development and validation of a quantitative image
analysis method to evaluate comet assay (silver
staining). J Bras Patol Med Lab.; 45 (4): 325-34.
Duarte, J.A., Furtado, A.P.A., Marroni, C.A., 2012. Use of
pineapple juice with gadopentetate dimeglumine as a
negative oral contrast for magnetic resonance
cholangiopancreatography: a multicentric study.
Abdominal Imaging; 37: 447- 56
Espinosa, M.G., Sosa, M., Rodríguez, L.M. De L.,
Córdova, T., Alvarado, J.B., Rodríguez, M.A.,
Aguilera, J.A.R., 2006. Blackberry (Rubus spp.): a pH-
dependent oral contrast medium for gastrointestinal
tract images by magnetic resonance imaging. Magnetic
Resonance Imaging; 24 195–200.
Fraga, T.C., Araujo, D.B., Sanchez, T.A., Elias, J.Jr.,
Carneiro, A.A.O., Oliveira R.B., Sosa, M., Baffa, O.,
2004. Euterpe olerácea (açaí) as an alternative oral
contrast agent in mri of the gastrointestinal system:
preliminar results. Magnetic Resonance Imaging.; 22;
389-93.
Ghanaati, H., Yazdi, H.R., Jallati, A.H., Abahashemi, F.,
Shakiba, M., Firouznia, K. 2011.Improvent of MR
cholangiopancreatography (MRCP) images after black
tea consumption. European Radiology, 21; 2551-57.
Griffin, N., Edwards, G.C., Grant, L.A., 2012. Magnetic
Resonance Cholangiopancreatography: the ABC of
MRCP. Insights Imaging; 3:11-21.
Image J. Online. http://imagej.nih.gov. Acessed 12
October. 2019.
Jornada, T.S., Murata, C.H., Medeiros, R.B., 2016.
Influence of partial k-space filing on the quality of
0
50
100
150
200
250
300
0 102030405060708090100
Gray levels
Width distance (mm)
Contrast A
Contrast B
1
2
4
3
BIOSIGNALS 2020 - 13th International Conference on Bio-inspired Systems and Signal Processing
104
magnetic resonance images. Radiologia Brasileira ; 49
(3): 158- 64.
Marques, P.M.A., Caritá, E.C., Benedicto, A.A., Sanches,
P.R.,
2005. Integração RIS/PACS no Hospital das
Clínicas de Ribeirão Preto: uma solução baseada em
web”. Radiologia Brasileira; 38 (1): 37- 43.
Mantau, J.N., Pinho, K.E.P., Costa, R.Z.V., Pinho, A.C.
Estudo comparativo do contraste negativo e contraste
natural em exame de colangiopancreatografia por
ressonância magnética. In: 44ª Jornada Paulista de
Radiologia [internet]. 2014, São Paulo, Painel
Impresso: Abdominal/Gastrointestinal. Available from:
˂http://static.spr.org.br/jpr/2014/trabalhosapresentados
/#modal-trabalho pdf˃
Pinho, K.E.P., Pinho, A.C., Gewehr, P.M., Gusso, A.M.
Image Quality in Magnetic Resonance Cholangi-
pancretography Exams: Study Between Açai Juice
and Manufactured Contrast Agent. In: Costa-Felix
R., Machado J., Alvarenga A. (eds) XXVI Brazilian
Congresso n Biomedical Engineering, 2018 IFMBE
Proceedings, 70/2. Springer, Singapore, Avaliable
from:˂ https://link.springer.com/chapter/10.1007/978-
981-13-2517-5_40˃ Acessed 24/10/2019
Pinho, K.E.P., Pinho, A.C., Pisani, J.C., Goedert, C. A.;
Magri, A.G., Gewehr, P.M. 2019.Açai juice as contrast
agent in MRCP exams: qualitative and quantitative
image evaluation. Brazilian Archives of Biology and
Technology .62 (e19160697), 1-14.
Pinho, K.E.P., Gewehr, P.M., Pinho, A.C., Pisani, J.C.
Sucos naturais como agentes de contraste para exames
de colangiopancreatografia por ressonância magnética.
In: XXIV Congresso de Engenharia
Biomédica[internet]. 2014; p. 1685-1688. Available
from:˂http://www.canal6.com.br/cbeb/2014/artigos/cb
eb2014_submission_499.pdf˃
Renzulli, M.; Biselli, M.; Fabbri, E.; Caretti, D.; Sergenti,
A.; Modestino, F.; Giannone, F.A.; Storchi, M.;
Pierotti, L.; Golfieri, R. 2019. What is the best fruit
juice to use a negative oral contrast agent in magnetic
resonance cholangiopancreatography? Clinical
Radiology, 74; 220-227.
Souza, V.B., Federizzi, M., Fernandes, D., Franco, A.R.:
Avaliação quantitativa da qualidade de exame de
imagem por ressonância magnética nuclear. In: XXIV
Congresso Brasileiro de Engenharia Biomédica
[internet]. 2014. Available from: http://www.canal6.
com.br/cbeb/2014/artigos/cbeb2014_submission_375.
pdf
Sanchez, T.A., Elias, J.Jr., Colnago, L.A., Troncon, L.E.A.,
Oliveira, R.B., Baffa, O., Araujo, D.B., 2009. Clinical
Feasibility of açaí (Euterpe olerácea) pulp as an oral
contrast agent for magnetic resonance
cholangiopancreatography. Journal of Computer
Assisted Tomography; 23(5), 666-70.
Xiao, B., Zhang, X.M., 2010. Magnetic resonance imaging
for acute pancreatitis. World Journal of Radiology;
August 28; 2(8): 298-308.
Westbrook, C., Roth, C.K.; Talbot, J.,2011. MRI in
Practice. 4
th
. Wiley Black Well.
Westbrook, C., 2010. Manual de Técnicas de Ressonância
Magnética. 3
a
ed. Rio de Janeiro: Guanabara Koogan.
Image Evaluation in Magnetic Resonance Cholangiopancreatography
105