DOI:10.2214/AJR.05.0522
AJR 2006; 187:464-472
© American Roentgen Ray Society
Double Hepatic Arterial Phase MRI of the Liver with Switching of Reversed Centric and Centric K-Space Reordering
Masayuki Kanematsu1,2,
Satoshi Goshima2,
Hiroshi Kondo2,
Ryujiro Yokoyama1,
Kimihiro Kajita1,
Hiroaki Hoshi2,
Minoru Onozuka3,
Atsushi Nozaki4,
Masaya Hirano4,
Yoshimune Shiratori5 and
Noriyuki Moriyama6
1 Department of Radiology Services, Gifu University Hospital, 1-1 Yanagido, Gifu
501-1193, Japan.
2 Department of Radiology, Gifu University School of Medicine, Gifu,
Japan.
3 Department of Physiology and Neuroscience, Kanagawa Dental College, Yokosuka,
Japan.
4 Imaging Application Technology Center, GE Yokogawa Medical Systems, Tokyo,
Japan.
5 Department of Medical Informatics, Gifu University School of Medicine, Gifu,
Japan.
6 Research Center for Cancer Prevention and Screening, National Cancer Center
Hospital, Tsukiji, Japan.
Received March 24, 2005;
accepted after revision June 7, 2005.
Supported in part by Health and Labor Sciences research grants for Third
Term Comprehensive Control Research for Cancer. Address correspondence to M.
Kanematsu.
Abstract
OBJECTIVE. The purpose of our study was to evaluate the clinical
feasibility and usefulness of a 2D spoiled gradient-recalled echo MR sequence
with serial switching of reversed centric and centric k-space reordering for
high-spatial-resolution gadolinium-enhanced double hepatic arterial phase
(HAP) MRI of the liver.
SUBJECTS AND METHODS. MR images (frequency, 512; phase encoding
without interpolation, 224; 6-mm thickness with 1-mm gap; 30 slices per 18
seconds) were obtained with multiphase imaging in which central k-space line
data were filled 10, 21, 49, and 181 seconds after arrival of contrast medium
in the abdominal aorta for the early HAP (reversed centric reordering, center
of k-space lines acquired at end of acquisition), late HAP (centric
reordering, center of k-space lines at beginning of acquisition), portal
venous phase (centric reordering), and equilibrium phase (centric reordering),
respectively, in 102 consecutive patients with suspected liver disease,
including 48 untreated hepatocellular carcinomas (HCCs) in 35 patients. Images
were quantitatively assessed for degree of contrast enhancement in the
abdominal aorta, spleen, portal trunk, liver parenchyma, hepatic veins, and
HCCs. Images were qualitatively assessed for the effectiveness of contrast
enhancement in each phase and for degree of image degradation due to
artifacts.
RESULTS. Enhancement of the abdominal aorta peaked in the early HAP,
of the portal trunk in the late HAP, and of the hepatic parenchyma and veins
in the portal venous phase. Mean HCC-to-liver contrast peaked in the early HAP
and turned to a negative value in the portal venous and equilibrium phases.
Sufficient image quality was achieved in 99 (97%) of the patients. One of the
other three patients had motion artifacts due to body motion, and the other
two had unsatisfactory respiratory suspension. Scan timing for early and late
HAP was optimal in 74 (73%) of the patients, for late HAP lagged in 20 (20%),
for early HAP was premature in six (6%), and for early HAP lagged in five (5%)
of the patients.
CONCLUSION. We confirmed the feasibility and usefulness of a 2D
gadolinium-enhanced double HAP spoiled gradient-recalled echo sequence
incorporating serial switching of reversed centric and centric k-space
reordering. This method has the potential for use in high-spatial-resolution
double HAP MRI for the diagnosis of hypervascular HCC.
Keywords: contrast media • high resolution • k-space reordering • liver • MR technique
Introduction
In recent years, the rapid scanning capability of MDCT has enabled multiple
scanning of the entire liver in a single hepatic arterial dominant phase of IV
contrast-enhanced CT. With this technique, radiologists can observe
hemodynamics in focal hepatic lesions and differentiate hypervascular
hepatocellular carcinoma (HCC) from benign tumors and pseudolesions
[1]. Meanwhile, a substantial
incremental gain in imaging speed has been achieved with the development of
MRI parallel acquisition
[2-4].
This technique has been applied to liver MRI to enable double hepatic arterial
phase (HAP) imaging of the liver after IV bolus injection of a gadolinium
chelate [5]. However, in double
HAP imaging of the liver during a single breath-hold, the numbers of imaging
matrices and obtainable slices must be reduced, even with the parallel
acquisition technique, because of the intrinsic trade-off with MRI parameters.
We needed an MRI sequence that maintained high spatial resolution while
enabling capture of efficient double HAP contrast with the use of
sophisticated k-space reordering strategies
[6]. We thus performed
whole-liver MRI with a 2D high-spatial-resolution gadolinium-enhanced double
HAP gradient-recalled echo sequence with serial switching and reversed centric
and centric k-space reordering. This technique enables acquisition of 30
6-mm-thick slices covering the entire liver at a frequency of 512 and
phase-encoding of 224 during an 18-second breath-hold. It also allows capture
of early and late HAPs with an interval of 11 seconds. This study was a
preliminary evaluation of the clinical feasibility and usefulness of double
HAP MRI of the liver.
View larger version (13K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1 —Drawing shows timing scheme of double hepatic arterial phase
(HAP) imaging. K-space lines for early HAP imaging are filled with echo data
from k-space margins to center, and those for other phase imaging are filled
from center to margins. Eight dummy excitation pulses are given for first
second. K-space centers are filled 10, 21, 49, and 181 seconds after arrival
of contrast medium in abdominal aorta. Practical imaging delay (D)
for early HAP imaging is determined as follows: D =
TV-A - 8, where TV-A is aortic transit
time in test bolus imaging. Eight seconds is subtracted so that end of first
HAP imaging (k-space center) comes 10 seconds after arrival of contrast medium
in abdominal aorta. Late HAP imaging begins automatically after 10-second
breathing interval after early HAP imaging. Portal venous phase imaging is
started 10 seconds after late HAP imaging. Equilibrium phase imaging is
initiated so that k-space lines are filled at 181 seconds.
|
|
Subjects and Methods
Patients
During the 3-month period from June to August 2003, 102 consecutive
patients with signs and symptoms of liver disease underwent
gadolinium-enhanced MRI of the liver in our department. These patients had
previously undergone sonography, CT, or laboratory evaluations. All patients
were informed that the radiologic examinations were primarily for clinical
diagnosis and secondarily for radiologic research. Thereafter, they all signed
waivers in conformity with the requirements of our institutional review
board.
These 102 patients, 75 men and 27 women (age range, 29-83 years; mean, 67.1
years), formed the study population. Forty-one of the 102 patients had HCC
with chronic hepatitis or cirrhosis, 24 had cirrhosis, 10 had chronic
hepatitis, five had metastatic lesions, three had cavernous hemangioma, one
had HCC without chronic hepatitis or cirrhosis, one had cholangiocellular
carcinoma, one had diffuse fatty liver, one had an inflammatory
myofibroblastic tumor, one had biliary hamartomatosis, and 14 had a healthy
liver.
MRI Protocol
MRI was performed with a 1.5-T superconducting MR system (Signa Excite Xl
TwinSpeed 1.5 T, GE Healthcare Technologies). The system provided a maximum
gradient strength of 23 mT/m, with a peak slew rate of 80 mT/m/ms for the
whole-body mode used in this study. All MR images were obtained in a
transaxial plane with an eight-channel phased-array multicoil and a field of
view of 38 cm. The MRI protocol consisted of dual-phase T1-weighted spoiled
gradient-recalled echo images (TR/TE, 245/4.2 in phase; 245/2.2 opposed phase;
320 x 224 frequency x phase encoding matrix; receiver bandwidth of
± 125 kHz; parallel imaging reduction factor of 2; one signal acquired;
and 27-second breath-hold acquisitions for 30 slices per breath-hold),
fat-suppressed respiration-triggered fast spin-echo T2-weighted images
(TR/effective TE, 7,500-9,200/80.6-88.6, echo-train length of 12-18, 512
x 256 matrix, receiver bandwidth of ± 62.5 kHz, reduction factor
of 2, 2 signals acquired, 3- to 4-minute acquisition time), and breath-hold
gadolinium-enhanced double HAP spoiled gradient-recalled echo images (155/1.5,
60° flip angle, 512 x 224 matrix without interpolation, 6-mm slice
thickness with 1-mm intersection gap, receiver band-width of ± 62.5
kHz, reduction factor of 2, one signal acquired, 30 slices per 18
seconds).
Test Bolus Imaging
Test bolus imaging was performed to determine the aortic transit time,
defined as the time from initiation of IV injection of contrast material to
peak enhancement in the abdominal aorta at the level of the first lumbar
vertebral body. Coronal single-section spoiled gradient-recalled echo images
(14.7/1.8, 30° flip angle, 256 x 128 matrix, receiver bandwidth of
± 31.3 kHz, 1 signal acquired, 1-second acquisition time) were obtained
every second after initiation of an IV bolus injection of 1 mL of
gadopentetate dimeglumine (Magnevist, Schering) followed by a flush with a
sterile saline solution. The volume of the flush was calculated from patient
body weight according to the following equation:
where V is the volume of saline solution in milliliters and
Wbody is patient body weight in kilograms. A power
injector (Sonic Shot 50, Nemotokyorindo) was used to inject first contrast
material and then the saline solution into an antecubital vein through a
22-gauge 25-mm-long catheter at a rate of 3 mL/s. Radiology technicians using
a circular cursor 15 mm in diameter obtained operator-defined
region-of-interest measurements of the mean signal intensity of the abdominal
aorta on test bolus gradient-recalled echo images. Aortic transit time in
seconds was determined as the time from initiation of contrast injection to
peak enhancement.
Double HAP MRI of the Liver
Breath-hold double HAP MRI of the liver was performed with a field of view
of 38 cm. MR images were obtained before and after IV bolus injection of 0.1
mmol of gadopentetate dimeglumine per kilogram of body weight followed by a
15-mL flush of sterile saline solution. The amount of sterile saline solution
flushed was fixed in all patients. The power injector used for test bolus
imaging was also used for injecting contrast material and saline solution at a
rate of 3 mL/s.
Figure 1 illustrates a
double HAP MRI timing scheme. The k-space lines in the phase-encoding
direction for early HAP imaging were filled with the center of k-space
performed at the end of scan acquisition (reversed centric reordering). The
next and subsequent acquisitions were performed with the center of k-space
acquired at the beginning of the sequences (centric reordering). Eight dummy
excitation pulses of approximately 1 second were given at the start of each
sequence. The center of the k-space lines was filled 10, 21, 49, and 181
seconds after the arrival of the contrast medium in the abdominal aorta. The
practical imaging delay (D, the time in seconds from initiation of
contrast injection to initiation of gradient-recalled echo image acquisition)
for early HAP imaging was determined in individual patients by use of the
following equation:
where TV-A is aortic transit time (from the antecubital
vein [V] to the abdominal aorta [A]) determined by means of
test bolus imaging. Eight seconds was subtracted from TV-A
so that the end of the first HAP imaging data acquisition occurred 10 seconds
after arrival of the contrast medium in the abdominal aorta, because the
central k-space lines determined the bulk of signal intensity on MR images.
Late HAP imaging was begun automatically after a 10-second breathing interval
after completion of early HAP imaging such that k-space lines were filled 21
seconds after arrival of the contrast medium in the abdominal aorta. Likewise,
portal venous phase imaging with centric k-space filling was started 10
seconds after the end of late HAP imaging such that k-space lines were filled
at 49 seconds. An equilibrium phase gradient-recalled echo sequence with
centric k-space filling was initiated 180 seconds after the initiation of
contrast injection such that k-space lines were filled at 181 seconds.
Quantitative Image Analysis
One radiologist obtained quantitative measurements on the images of 97
patients. The quality of the images of three of the five excluded patients was
substantially impaired by motion artifacts, and the other two patients had
advanced multiple liver tumors. Operator-defined region-of-interest
measurements were obtained on unenhanced, early HAP, late HAP, portal venous
phase, and equilibrium phase images. Mean signal intensity was measured in the
abdominal aorta, spleen, portal trunk, right and left lobes of the liver, and
hepatic veins. A circular cursor 5-40 mm in diameter was used to measure the
signal intensity of anatomic structures. Signal intensity in the liver and
spleen was measured in regions without large vessels, dilated intra-hepatic
biliary ducts, or prominent artifacts. Because of the near-field effect of the
surface coil, cursors were placed so that slice level and distance from the
abdominal walls were consistent for the different imaging sequences. In each
patient, two hepatic parenchymal (right and left lobes) and two hepatic venous
(right or middle hepatic and left hepatic veins) measurements were averaged to
obtain the mean signal intensity for even sampling of signal intensity in the
liver. Spleen-to-liver contrast was calculated by subtracting the mean signal
intensity of the liver from that of the spleen.
In 48 untreated hypervascular HCC nodules ranging in size from 8 to 83 mm
(mean, 22 mm) in 35 patients who were 48-83 years old (mean, 71 years), signal
intensity was measured in the HCC and in the surrounding liver. Proof of HCC
was obtained with percutaneous liver biopsy (n = 15) or a
substantially increased 
-fetoprotein level or level of protein induced
by vitamin K absence or antagonism (PIVKA) II in combination with follow-up CT
or MRI showing tumor progression (n = 33). Tumor-to-liver contrast
was calculated by subtracting the mean signal intensity of an HCC from that of
the surrounding liver.
Qualitative Image Analysis
Two gastrointestinal radiologists independently reviewed the early HAP,
late HAP, portal venous phase, and equilibrium phase images with reference to
the unenhanced images and then reached consensus by discussion. The images
were evaluated on a radiologic viewer by means of subjective assessment by the
reviewers, who were blinded to patient clinical information.
We defined the characteristics of early HAP as spleen intensely enhanced,
proximal portal veins and liver parenchyma not or minimally enhanced, and
hepatic veins not enhanced; of late HAP as spleen intensely enhanced, proximal
portal veins and liver parenchyma minimally to moderately enhanced, and
hepatic veins not or minimally enhanced; of the portal venous phase as liver
parenchyma enhanced as intensely as the spleen; and of the equilibrium phase
as all vessels weakly enhanced to the same degree. On the basis of these
definitions, the reviewers recorded their impression for the overall image
value of each contrast phase as good, fair, or poor. The reviewers also
recorded the pertinence of scan timing for each contrast phase as premature,
optimal, or lagged. Finally, the reviewers assessed overall image degradation
due to artifacts (i.e., motion, susceptibility, or blurring) as none, mild,
moderate, or severe. "Severe" was assigned when the image
interpretation was hampered because of image degradation.
For the 48 untreated hypervascular HCC nodules, the reviewers used a
seven-point scale to evaluate the degree of hepatic arterial enhancement or
washout of HCC in each phase. A score of 3 was assigned for intense
enhancement, 2 for moderate enhancement, 1 for mild enhancement, 0 for no
enhancement or washout, -1 for mild washout, -2 for moderate washout, and -3
for distinct washout.
Statistical Analysis
Repeated measures analysis of variance and multiple comparisons with the
Scheffé criterion [7]
were performed to assess the transition of mean signal intensities in the
abdominal aorta, spleen, portal trunk, liver parenchyma, and hepatic veins and
spleen-to-liver and tumor-to-liver contrast. The Kruskal-Wallis test and
multiple comparisons with the Scheffé criterion were performed for
evaluation of qualitative degrees obtained as categoric data. Interobserver
variability was assessed with the kappa test. A kappa value up to 0.20 showed
slight agreement, a value of 0.21-0.40 showed fair agreement, 0.41-0.60 showed
moderate agreement, 0.61-0.80 showed substantial agreement, and 0.81 or
greater showed almost perfect agreement.
Results
The signal intensity of the abdominal aorta peaked in the early HAP and
then decreased over time. The signal intensity of the spleen was high in the
early HAP, late HAP, and portal venous phase and decreased slightly in the
equilibrium phase (Fig. 2). The
signal intensity of the portal trunk increased steeply over the late HAP and
decreased gradually. The signal intensity of liver parenchyma constantly
increased over the portal venous phase and then decreased slightly in the
equilibrium phase. The signal intensity of hepatic veins increased steeply
over the portal venous phase and then decreased slightly in the equilibrium
phase (Fig. 3). Spleen-to-liver
contrast peaked in the early HAP and then decreased steadily with time.
Tumor-to-liver contrast peaked in the early HAP and then constantly decreased
with time and was negative during the portal venous and equilibrium phases
(Fig. 4).
View larger version (12K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2 —Graph shows contrast phase versus mean signal intensity for
abdominal aorta and spleen. Signal intensity of abdominal aorta peaks in early
hepatic arterial phase (HAP) and then decreases over time. Significant
differences in mean signal intensity (p < 0.005) exist between all
phases but not between late HAP and portal venous phase (PVP). Signal
intensity of spleen is high in early HAP, late HAP, and portal venous phase
and then decreases slightly in equilibrium phase (EP). Significant differences
in mean signal intensity (p < 0.0001) exist between all phases but
not between early HAP and equilibrium phase or between late HAP and portal
venous phase. Pre = before contrast injection.
|
|
View larger version (14K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3 —Graph shows contrast phase versus mean signal intensity for
portal trunk, liver parenchyma, and hepatic veins. Signal intensity of portal
trunk increases steeply over late hepatic arterial phase (HAP) and then
decreases gradually. Significant differences in mean signal intensity
(p < 0.05) exist between all phases but not between late HAP and
portal venous phase (PVP). Signal intensity of liver parenchyma increases
constantly over portal venous phase and then decreases slightly in equilibrium
phase (EP). Significant differences in mean signal intensity (p <
0.05) exist between all phases but not between late HAP and equilibrium phase
or between portal venous and equilibrium phases. Signal intensity of hepatic
veins increases steeply over portal venous phase and then decreases slightly
in equilibrium phase. Significant differences in mean signal intensity exist
between all phases (p < 0.005). Pre = before contrast
injection.
|
|
View larger version (10K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4 —Graph shows contrast phase versus tumor-to-liver contrast.
Tumor-to-liver contrast peaks in early hepatic arterial phase (HAP) and then
decreases rapidly. This value turns negative during portal venous phase (PVP)
and equilibrium phase (EP). Significant differences in mean tumor-to-liver
contrast (p < 0.005) exist between all phases but not between
unenhanced and portal venous phases, unenhanced and equilibrium phases, or
portal venous and equilibrium phases. Pre = before contrast injection.
|
|
The value of early or late HAP images was good or fair in most patients
(Table 1). The scan timing of
early HAP and late HAP was optimal in 74 (73%) of the 102 patients, of late
HAP lagged in 20 (20%), of early HAP was premature in six (6%), and of early
HAP lagged in five (5%) of the patients. Scan timing for the portal venous
phase and the equilibrium phase was optimal in all patients
(Table 2). Sufficient image
quality for image interpretation was achieved in 99 (97%) of the patients
(Figs. 5A,
5B,
5C,
5D,
5E,
6A,
6B,
6C, and
6D). One of the other three
patients had motion artifacts due to body motion, and two had unsatisfactory
respiratory suspension (Table
3).
View this table:
[in this window]
[in a new window]
|
TABLE 2: Pertinence of Scan Timing for Early and Late Hepatic Arterial Phases
(HAPs) Qualitatively Assessed by Two Radiologists
|
|
View larger version (147K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A —63-year-old man with developing hypervascular hepatocellular
carcinoma (HCC) and cirrhosis due to type C viral hepatitis. Unenhanced
spoiled gradient-recalled echo axial image (TR/TE, 155/1.5) shows hepatic
nodule (arrow) with 3-cm area of mixed signal intensity in posterior
segment of liver.
|
|
View larger version (146K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B —63-year-old man with developing hypervascular hepatocellular
carcinoma (HCC) and cirrhosis due to type C viral hepatitis. Early hepatic
arterial phase (HAP) spoiled gradient-recalled echo axial image (155/1.5)
obtained with reversed centric k-space reordering and for which central
k-space lines were filled 10 seconds after arrival of contrast medium in
abdominal aorta shows intensely enhanced abdominal aorta (asterisk)
and proximal hepatic arteries (arrowheads), splenic moiré
pattern enhancement (curved arrow), and HCC as area of homogeneous
enhancement (straight arrow).
|
|
View larger version (149K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5C —63-year-old man with developing hypervascular hepatocellular
carcinoma (HCC) and cirrhosis due to type C viral hepatitis. Late HAP spoiled
gradient-recalled echo axial image (155/1.5) obtained with centric k-space
reordering and for which central k-space lines were filled 21 seconds after
arrival of contrast medium in abdominal aorta shows intensely enhanced
proximal portal veins (arrowheads) and HCC as mixed area of
persistent enhancement (arrow) and washout. Splenic enhancement is
more intense than hepatic parenchymal enhancement.
|
|
View larger version (151K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5D —63-year-old man with developing hypervascular hepatocellular
carcinoma (HCC) and cirrhosis due to type C viral hepatitis. Portal venous
phase spoiled gradient-recalled echo axial image (155/1.5) obtained with
centric k-space reordering and for which central k-space lines were filled 49
seconds after arrival of contrast medium in abdominal aorta shows HCC as mixed
area of mild washout (arrow). Hepatic parenchymal enhancement is as
intense as splenic enhancement.
|
|
View larger version (150K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5E —63-year-old man with developing hypervascular hepatocellular
carcinoma (HCC) and cirrhosis due to type C viral hepatitis. Equilibrium phase
spoiled gradient-recalled echo axial image (155/1.5) obtained with centric
k-space reordering and for which central k-space lines were filled at 181
seconds after contrast arrival in abdominal aorta shows HCC as area of clear
washout (arrow).
|
|
View larger version (123K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6A —64-year-old man with surgically proven 12-mm hypervascular
hepatocellular carcinoma (HCC) and cirrhosis due to type B viral hepatitis.
Early hepatic arterial phase (HAP) spoiled gradient-recalled echo axial image
(TR/TE, 155/1.5) obtained with reversed centric k-space reordering and for
which central k-space lines were filled 10 seconds after arrival of contrast
medium in abdominal aorta shows HCC as area of homogeneous enhancement
(arrow).
|
|
View larger version (124K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6B —64-year-old man with surgically proven 12-mm hypervascular
hepatocellular carcinoma (HCC) and cirrhosis due to type B viral hepatitis.
Late HAP spoiled gradient-recalled echo axial image (155/1.5) obtained with
centric k-space reordering and for which central k-space lines were filled 21
seconds after arrival of contrast medium in abdominal aorta shows HCC as area
of ringlike coronal enhancement (arrow).
|
|
View larger version (123K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6C —64-year-old man with surgically proven 12-mm hypervascular
hepatocellular carcinoma (HCC) and cirrhosis due to type B viral hepatitis.
Portal venous phase spoiled gradient-recalled echo axial image (155/1.5)
obtained using centric k-space reordering and for which central k-space lines
were filled 49 seconds after arrival of contrast medium in abdominal aorta
shows HCC as area of subtly persistent coronal enhancement
(arrow).
|
|
View larger version (123K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6D —64-year-old man with surgically proven 12-mm hypervascular
hepatocellular carcinoma (HCC) and cirrhosis due to type B viral hepatitis.
Equilibrium phase spoiled gradient-recalled echo axial image (155/1.5)
obtained with centric k-space reordering and for which central k-space lines
were filled 181 seconds after arrival of contrast medium in abdominal aorta
shows almost no abnormal imaging findings for HCC. Ringlike coronal
enhancement in B is crucial in differential diagnosis between
hypervascular HCC and early enhancing pseudolesion, because equilibrium phase
image shows no abnormal imaging findings such as tumor washout or delayed
enhancement of fibrous pseudocapsules.
|
|
The qualitative results corresponded well to the quantitative results shown
in Figure 4. In three (9%) of
the 35 patients with HCC, tumors not depicted in the early HAP because of
premature scan timing were well depicted as hypervascular lesions in the late
HAP (Fig. 7). Kappa values for
the two reviewers who performed the independent rating of images were found to
range from 0.71 to 1.00 (mean, 0.80), indicating substantial to almost
complete agreement.
View larger version (8K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7 —Graph shows contrast phase versus degree of hepatic arterial
enhancement or washout of hepatocellular carcinoma. Degree peaked in early
hepatic arterial phase (HAP) and then decreased rapidly. Degree turned
negative over portal venous phase (PVP) and equilibrium phase (EP).
Significant differences in mean degree (p < 0.05) exist between
all phases. Qualitative results correspond well with quantitative results in
Figure 4.
|
|
Discussion
Double HAP contrast-enhanced CT, which allows acquisition of both early and
late arterial hepatic images, reveals hemodynamic characteristics of malignant
hepatic tumors. Moreover, interpretation of images obtained during both
arterial phases has shown that findings on images in these phases have the
greatest sensitivity and positive predictive value
[1]. MRI offers information on
T1 and T2 relaxation times with unenhanced MR images, on fat deposition with
phase-shift gradient-recalled echo images, and on hemodynamics with double HAP
imaging. There are, however, intrinsic trade-offs in MRI with respect to
imaging matrices and number of obtainable slices that apply to our double HAP
sequence. Our MRI sequence maintained large imaging matrices of 512 frequency
and 224 phase encoding without zero-fill interpolation and allowed acquisition
of 30 6-mm-thick images by extending the acquisition time for a single phase
to 18 seconds. The rationale behind this imaging sequence was that it enables
capture of double HAP contrast over a 10-second breathing interval by
switching the direction of k-space reordering from reversed centric to centric
filling during the early and late HAPs.
The series of scan times we set for multiphase imaging was optimal in most
patients in our study. Scan timing for the early HAP was premature in 6% of
the patients, although we used test bolus imaging before liver imaging in all
patients. Even in such cases, late HAP imaging was helpful in the detection of
hypervascular HCCs that were missed in the early HAP. We may, however, need to
push the scan timing of the late HAP ahead somewhat because it lagged in 20%
of the patients in our study. Regarding image quality, the spoiled
gradient-recalled echo sequence we used was robust for the use of double HAP
imaging, as the qualitative assessment of image degradation indicated.
Some researchers have reported on the usefulness of the 3D
gradient-recalled echo sequence in gadolinium-enhanced MRI of the liver
[8], although few reports on
the application of a 3D gradient-recalled echo sequence for double HAP imaging
are available. Although a 3D sequence was initially developed for use in
gadolinium-enhanced MR angiography, this sequence has been widely applied to
contrast-enhanced MRI of the liver. Our present technique with a 2D sequence
may apply for double HAP imaging with a 3D sequence.
Ueda et al. [9], who
investigated the hemodynamics of hypervascular HCC with single-level dynamic
CT during hepatic arteriography, reported that whole HCC enhancement and clear
contour visualization began 4.5 seconds after the start of hepatic arterial
contrast injection; contrast enhancement in the adjacent liver began at 10.5
seconds; and the contrast material was washed out from the tumor and coronal
enhancement of adjacent liver appeared at 22.4 seconds. On the basis of their
results, we set a breathing interval between the early and late HAPs such that
the filling of each central k-space occurred at intervals of 11 seconds.
Eight-channel coil technology contributed to the efficiency of our
sequence. Multichannel phased-array coils offer increased signal-to-noise
ratio over standard coils near the array elements while preserving
signal-to-noise ratio at the center of the object
[10,
11]. Despite the use of large
imaging matrices, a thin slice, and the parallel acquisition technique, we
were able to maintain sufficient image contrast with our sequence because of
the improved signal-to-noise capability of eight-channel phased-array
coils.
Regarding reduction factor number for the parallel imaging technique, image
acquisition accelerated by a factor of 2-3 is common for many clinical
applications [12]. We used a
reduction factor of 2 for our sequence. Although this number could have been
set greater than 2 to further shorten the acquisition time, reduced
signal-to-noise ratio and increased risk of wrap-around artifact became
problematic in a preliminary evaluation.
There are possible drawbacks to the double HAP sequence. First, image
sharpness of the first HAP images might have been impaired because both top
and bottom peripheral k-space lines filled before sufficient contrast material
had reached the liver or tumors. Nevertheless, there was no significant image
blurring, and hypervascular HCCs were well depicted on the first HAP images.
We infer that this blurring did not affect clinical image interpretation.
Second, slice levels might have been staggered on the early and late HAP
images because there was an in-between breathing interval. This drawback was
compensated by the thin slices used.
There were limitations to our study. First, there was lack of pathologic
standard of reference in many cases, and we did not assess observer
performance in terms of the diagnosis of focal liver lesions. The real effect
of the sequence on clinical radiologic practice should be clarified by results
of future biostatistical studies. Although we set an interval of 11 seconds
between the central k-space lines in the early HAP and those in the late HAP,
this interval may not necessarily be optimal in light of the hemodynamics of
hypervascular tumors. The interval requires optimization. Finally, because we
used the parallel imaging technique, we could not normalize signal-to-noise
and contrast-to-noise ratios using SDs of the signal intensities of
background, which were measured in the phase-encoding direction outside the
anterior abdominal wall. Instead, we used mean signal intensity measured by
the region-of-interest method to assess transition of degree of contrast
enhancement over multiphase imaging. We assumed that background noise was
constant throughout the multiple contrast phases.
In conclusion, in whole-liver imaging we used a 2D gadolinium-enhanced
double HAP spoiled gradient-recalled echo imaging sequence incorporating
serial switching of reversed centric and centric k-space reordering. This
technique has the potential for use as a high-spatial-resolution
gadolinium-enhanced double HAP MRI sequence for the diagnosis of hypervascular
HCC.
References
- Murakami T, Kim T, Takamura M, et al. Hypervascular hepatocellular
carcinoma: detection with double arterial phase multi-detector row helical CT.
Radiology 2001;218
: 763-767[Abstract/Free Full Text]
- Carlson JW. An algorithm for NMR imaging reconstruction based on
multiple RF receiver coils. J Magn Reson1987; 74:376
-380
- Carlson JW, Minemura T. Imaging time reduction through multiple
receiver coil data acquisition and image reconstruction. Magn Reson
Med 1993; 29:681
-688[Medline]
- Sodickson DK, Manning WJ. Simultaneous acquisition of spatial
harmonics (SMASH): fast imaging with radio-frequency coil arrays.
Magn Reson Med 1997;38
: 591-603[Medline]
- Yoshioka H, Sato J, Takahashi N, et al. Dual double arterial phase
dynamic MR imaging with sensitivity encoding (SENSE): which is better for
diagnosing hypervascular hepatocellular carcinomas, in-phase or opposed-phase
imaging? Magn Reson Imaging 2004;22
: 361-367[CrossRef][Medline]
- Maki JH, Prince MR, Londy FJ, Chenevert TL. The effects of time
varying intravascular signal intensity and k-space acquisition order on
three-dimensional MR angiography image quality. J Magn Reson
Imaging 1996; 6:642
-651[Medline]
- Fleiss JL, ed. The analysis of variance and multiple
comparisons: the design and analysis of clinical experiments. New
York, NY: Wiley, 1986: 51-59
- Coulam CH, Chan FP, Li KC. Can a multiphasic contrast-enhanced
three-dimensional fast spoiled gradient-recalled echo sequence be sufficient
for liver MR imaging? AJR 2002;178
: 335-341[Abstract/Free Full Text]
- Ueda K, Matsui O, Kawamori Y, et al. Hypervascular hepatocellular
carcinoma: evaluation of hemodynamics with dynamic CT during hepatic
arteriography. Radiology 1998;206
: 161-166[Abstract/Free Full Text]
- Porter JR, Wright SM, Reykowski A. A 16-element phased-array head
coil. Magn Reson Med 1998;40
: 272-279[Medline]
- Ozsarlak O, Van Goethem JW, Parizel PM. 3D time-of-flight MR
angiography of the intracranial vessels: optimization of the technique with
water excitation, parallel acquisition, eight-channel phased-array head coil
and low-dose contrast administration. Eur Radiol2004; 14:2067
-2071[CrossRef][Medline]
- Heidemann RM, Ozsarlak O, Parizel PM, et al. A brief review of
parallel magnetic resonance imaging. Eur Radiol2003; 13:2323
-2337[CrossRef][Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
S. Goshima, M. Kanematsu, H. Kondo, Y. Shiratori, M. Onozuka, N. Moriyama, and K. T. Bae
Optimal Acquisition Delay for Dynamic Contrast-Enhanced MRI of Hypervascular Hepatocellular Carcinoma
Am. J. Roentgenol.,
March 1, 2009;
192(3):
686 - 692.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
