DOI:10.2214/AJR.05.0706
AJR 2006; 187:W255-W266
© American Roentgen Ray Society
Optimal TE for SPIO-Enhanced Gradient-Recalled Echo MRI for the Detection of Focal Hepatic Lesions
Myeong-Jin Kim1,2,3,
Joo Hee Kim1,3,
Jin Young Choi1,
Sung Ho Park1,
Jae-Joon Chung1,4,
Ki Whang Kim1,3 and
Donald G. Mitchell5
1 Department of Diagnostic Radiology, Yonsei University College of Medicine,
Seodaemun-ku Shinchon-dong 134, Seoul 120-752, Republic of Korea.
2 Brain Korea 21 Project for Medical Science and Institute of Gastroenterology,
Severance Hospital, Yonsei University College of Medicine, Seoul, Republic of
Korea.
3 Institute of Radiological Science, Severance Hospital, Yonsei University
College of Medicine, Seoul, Republic of Korea.
4 Department of Diagnostic Radiology, NHIC Ilsan Hospital, Gyonggi-do,
Korea.
5 Department of Radiology, Thomas Jefferson University Hospital, Philadelphia,
PA.
Received April 26, 2005;
accepted after revision July 10, 2005.
Address correspondence to M.-J. Kim
(kimnex{at}yumc.yonsei.ac.kr).
Supported by Yonsei University Research Fund of 2006.
WEB
This is a Web exclusive article.
Abstract
OBJECTIVE. The objective of our study was to determine the optimal
TE for superparamagnetic iron oxide (SPIO)-enhanced gradient-recalled echo
(GRE) MRI for the detection of focal hepatic lesions.
MATERIALS AND METHODS. Ferucarbotran-enhanced GRE sequences,
performed on a 1.5-T MR system, were used to evaluate 131 lesions (38
hepatocellular carcinomas, 37 metastases, 21 hemangiomas, 7
cholangiocarcinomas, 15 cysts, and 13 miscellaneous lesions) at four different
TEs: 9, 13.5, 18, and 22.5 milliseconds. The lesion-to-liver signal
difference-to-noise ratio (SDNR) was compared among the four GRE sequences by
paired Student's t tests and among lesion types by an independent
samples Student's t test. The McNemar test was used to compare the
sensitivity for the detection of focal hepatic lesions. Wilcoxon's signed rank
test was used to compare the subjective lesion conspicuity.
RESULTS. The SDNRs of lesions on GRE images obtained at a TE of 13.5
milliseconds (mean ± SD, 60 ± 24) were significantly (p
< 0.001) higher than those at TEs of 9 (55 ± 23), 18 (55 ±
22), and 22.5 milliseconds (47 ± 19). The SDNR was highest at a TE of
13.5 milliseconds for SPIO-uptake lesions and was comparable on images
obtained with TEs of 18 and 13.5 milliseconds for non-SPIO-uptake lesions. The
non-SPIO-uptake lesions showed a significantly higher SDNR than the
SPIO-uptake lesions at a TE of 22.5 milliseconds (p = 0.007). The
overall sensitivity for lesion detection was not significantly different among
the four GRE sequences, and the subjective ratings of lesion conspicuity were
comparable for images obtained using TEs of 8, 13.5, and 18 milliseconds, but
the ratings of lesion conspicuity were significantly lower for images obtained
using a TE of 22.5 milliseconds (p < 0.001).
CONCLUSION. For ferucarbotran-enhanced MRI, lesion SDNR was highest
on images obtained using a TE of 13.5 milliseconds, but the sensitivity and
lesion conspicuity were comparable at TEs of 9 and 18 milliseconds. The SDNR
of liver lesions varied according to the lesion's potential capability of
taking up SPIO agents.
Keywords: contrast media • liver • liver disease • MR contrast agents • MR technique
Introduction
Superparamagnetic iron oxide (SPIO) agents are being used for hepatic MRI
examination to improve the detection and characterization of focal hepatic
lesions: The use of ferumoxides has gained worldwide approval, and
ferucarbotran has been approved for use in Europe and Asia
[1-3].
Although various pulse sequences, including fast spin-echo and
gradient-recalled echo (GRE) images, are adapted by many investigators for
SPIO-enhanced MRI, T2*-weighted GRE sequences are regarded as an
essential technique for the detection of focal hepatic lesions
[4-7].
However, scanning parameters have usually been determined empirically
[1]. Among the various scanning
parameters, the contrast enhancement effect of SPIO is most dependent on the
TE because the magnetic susceptibility increases as the TE is increased.
In prior reports, SPIO-enhanced GRE images were obtained at TE values that
varied from 2.1 to 22.5 milliseconds, depending on the investigation
[6-11].
To define the optimal TE values for ferumoxides-enhanced GRE images, Matsuo et
al. [12] recently compared GRE
images obtained at five different TEs: 1.4, 4.2, 6, 8, and 10 milliseconds. In
that study, GRE images obtained at a TE of 8 milliseconds performed better in
terms of enabling detection of malignant lesions than those obtained at
shorter TEs and a TE of 10 milliseconds. In a study by Ward et al.
[13], GRE images obtained
using ferumoxides at a TE of 15 milliseconds were more sensitive for the
detection of colorectal metastasis than GRE images obtained at a TE of 11
milliseconds and fast spin-echo images. However, the two studies compared
different ranges of TE values, and it is still not known whether GRE images
obtained with a TE of longer than 15 milliseconds will allow improved lesion
detection. Furthermore, whether lesion conspicuity on GRE images depends on
the lesion character has not been fully evaluated. We conducted this study to
determine the optimal TE values in SPIO-enhanced GRE MRI for the depiction of
various focal hepatic lesions on a 1.5-T system.
Materials and Methods
Patient Population
Between December 2003 and August 2004, contrast-enhanced GRE images were
obtained using four different TEs (9, 13.5, 18, and 22.5 milliseconds) in 71
consecutive patients (44 men and 27 women; age range, 28-75 years [mean, 56
years]; body weight range, 48-100 kg [mean, 65 kg]) who underwent
SPIO-enhanced hepatic MRI examinations. These examinations were performed
because the findings of prior sonography, helical CT, or gadolinium-enhanced
MRI were inconclusive or because further clarification about the number of
lesions seen was needed. This study followed the guidelines of the
institutional review board at our hospital, and informed written consent was
obtained from all of the patients.
One hundred thirty-one focal hepatic lesions were confirmed either
histologically or clinically. There were 38 hepatocellular carcinomas (HCCs)
in 25 patients, 37 metastases in 17 patients, 21 hemangiomas in 16 patients,
15 cysts in 12 patients, 8 focal eosinophilic infiltrations in 3 patients, 7
intrahepatic cholangiocarcinomas in 5 patients, and 5 epithelioid
hemangioendotheliomas in 1 patient. Eight patients had two types of lesion: 4
patients had HCC and cysts, 1 patient had a metastasis and 2 cysts, 1 patient
had a metastasis and a hemangioma, and 2 patients had hemangiomas and
cysts.
Liver metastases arose from the following primary tumors: colorectal
carcinoma (15 lesions in 7 patients), gastric carcinoma (7 lesions in 3
patients), bronchogenic carcinoma (3 lesions in 2 patients), bile duct
carcinoma (6 lesions in 2 patients), duodenal gastrointestinal stromal tumor
(5 lesions in 1 patient), and breast carcinoma (1 lesion in 1 patient). The
diameters of all the lesions ranged from 0.4 to 10.7 cm (mean, 2.5 cm).
Specifically, the diameters of intrahepatic cholangiocarcinomas ranged from
1.0 to 10.7 cm (mean, 5.7 cm); cysts, from 0.8 to 1.8 cm (mean, 1.3 cm);
epithelioid hemangioendotheliomas, from 1.6 to 2.4 cm (mean, 2.0 cm); focal
eosinophilic infiltrations, from 1.1 to 2.9 cm (mean, 2.1 cm); HCCs, from 0.4
to 12.0 cm (mean, 2.5 cm); hemangiomas, from 0.4 to 4.1 cm (mean, 1.4 cm); and
metastases, from 0.5 to 5.2 cm (mean, 2.1 cm).
Diagnosis was established as follows: A diagnosis of HCC was based on the
findings of a percutaneous biopsy (2 lesions in 2 patients); on findings in a
surgical specimen (7 lesions in 5 patients); or on typical clinical and
laboratory findings in combination with the progression of the disease, as
depicted on follow-up CT or MR images (29 lesions in 18 patients). In patients
with metastasis to the liver, diagnosis was based on histologic findings of
the primary tumor and rapid disease progression as depicted by serial
follow-up images (27 lesions in 9 patients) or on surgery (10 lesions in 8
patients). Hemangiomas were diagnosed on the basis of their typical appearance
on either sonography or CT and the absence of growth during a follow-up period
of at least 6 months (19 lesions in 14 patients), on gross examination during
surgery (1 lesion in 1 patient), or by surgical resection (1 lesion in 1
patient). Cysts were confirmed by typical findings on unenhanced T1- and
T2-weighted MR images, and the MRI findings were correlated with either
sonography or CT findings. Focal eosinophilic infiltrations were confirmed on
the basis of biopsy results and the appropriate clinical findings.
Intrahepatic cholangiocarcinomas were diagnosed on the basis of findings at
surgery (n = 1) or biopsy (n = 4). Finally, diagnosis of
epithelioid hemangioendothelioma was based on biopsy results.
Liver cirrhosis was diagnosed in 24 patients on the basis of clinical
findings (n = 19) or surgery or biopsy findings (n = 5). The
causes of liver cirrhosis were chronic hepatitis B virus in 20 patients,
hepatitis C virus in 2, alcoholism in 1, and cryptogenic in 1. Three patients
had hepatitis B virus without cirrhosis and 1 had hepatitis C virus without
cirrhosis. Among the 38 HCCs, 29 (76%) were found in patients with hepatitis B
virus; 4, in patients with hepatitis C virus; 1, in a patient with alcoholic
cirrhosis; and 1 in a patient with cryptogenic cirrhosis. Seventeen
hemangiomas (81%) were found in noncirrhotic patients and 4 in cirrhotic
patients. Nine cysts (60%) were found in noncirrhotic patients. All other
lesions were found in noncirrhotic patients.
MRI
MRI was performed on 1.5-T units (Signa Horizon, GE Healthcare). All images
were obtained in the transverse plane using a phased-array multicoil. A
rectangular field of view was adjusted for each patient in the range of 22-24
x 29-34 cm and was held constant for all sequences for each patient.
Unenhanced MRI was performed using four sequences. One was a
respiratory-triggered T2-weighted fast spin-echo sequence with a TR range of
3,500-10,900 milliseconds and effective TE range of 96-105 milliseconds,
echo-train length of 16, 2 signals acquired, a matrix of 256 x 256,
superior and inferior spatial presaturation and chemically selective fat
saturation, and a 7-to 8-mm slice thickness with a 1- to 2-mm gap. Two
additional sequences were a breath-hold T1-weighted fast multiplanar spoiled
GRE in-phase sequence (TR range/TE range, 150-200/4.2-4.6) followed by an
out-of-phase sequence (120-180/1.5-2.3) with a flip angle of 90°, 1 signal
acquired, a matrix of 256128, a 7- to 8-mm section thickness, and zero gap in
1 or 2 acquisitions. Breath-hold T2-weighted single-shot half-Fourier images
were obtained with an effective TE of 180 milliseconds, a matrix of 256
x 160, and a 7- to 8-mm section thickness with a 1- to 2-mm gap for the
fourth sequence.
For SPIO-enhanced MRI, a fixed volume of 1.4 mL (7.0-14.5 µmol of iron
per kilogram of body weight) of ferucarbotran (Resovist, Schering) was
administered as a rapid bolus by the hand injection method and was immediately
followed by a saline solution flush of 15-20 mL at a rate of approximately 2-3
mL/s. Dynamic MRI was performed using a chemically selective fat-suppressed
spoiled GRE sequence in the transverse plane during suspended respiration
immediately after the IV injection of the ferucarbotran. Additional images
were obtained at 30-35 seconds, 65-70 seconds, and 5 minutes after the
injection. The imaging parameters included TR range/TE range, 180-200/1.5-2.2;
90° flip angle; 256 x 128 matrix with a three quarters rectangular
field of view; 1 signal acquired; and 8- to 10-mm-thick sections with a zero
intersection gap.
After the ferucarbotran-enhanced dynamic and 5-min delayed images were
obtained, accumulation phase images were obtained 10 minutes after the
administration of the contrast agent. Four sets of the breath-hold
T2*-weighted unspoiled steady-state GRE images were obtained with a
TR of 200 milliseconds, flip angle of 30°, slice thickness of 7-8 mm with
zero gap, receiver bandwidth of 32 kHz, and matrix of 256 x 160 at
different TEs in the following order: 22.5, 18, 13.5, and 9 milliseconds. Each
TE value was chosen to obtain the images in phase and to encompass the range
of TEs used in previous studies but that had not been directly compared
[5,
7,
12-14].
Each GRE sequence was designated as GRE22.5, GRE18,
GRE13.5, and GRE9, respectively. In most patients, the
time for a single breath-hold was 27 seconds. Using that breath-holding time,
the number of slices per acquisition was 6, 7, 9, and 11 slices, respectively.
Therefore, to overlap the end slice of each acquisition, 5, 4, 3, and 2
acquisitions were needed in each sequence. All the SPIO-enhanced GRE images
were obtained between 10 and 30 minutes after the injection of the SPIO
agent.
Image Analysis
For this study, the four sets of ferucarbotran-enhanced GRE images, but not
the dynamic images, were evaluated. One observer drew a region of interest
(ROI) in the focal lesion and hepatic parenchyma to measure the signal
intensity (SI) of the focal lesions, hepatic parenchyma, and background noise
at a workstation (Centricity, GE Healthcare) using standard software. To
minimize the effects of partial volume averaging, only 106 lesions with
diameters greater than 1 cm were included in this analysis: 35 HCCs, 27
metastases, 19 hemangiomas, 13 cysts, 3 focal eosinophilic infiltrations, 6
intrahepatic cholangiocarcinomas, and 3 epithelioid hemangioendotheliomas. The
area of the ROI in the tumor was set to measure the homogeneous area of the
lesion while avoiding areas of necrosis, hemorrhage, or fibrosis and
attempting to maintain at least a 20-mm2 ROI when possible. The ROI
of the liver was drawn in the hepatic parenchyma near the tumor, to exclude a
vessel and ghosting artifacts, and was at least 80 mm2 in all
cases.
The SI of the background noise was measured in a circular area that was
ventral to the liver and outside the patient along the phase-encoding
direction. The circular area also included ghosting artifacts. In some cases,
respiratory ghosting artifacts were seen only at one or two particular
sequences because the patients failed to hold their breath during these
breath-hold sequences. In these cases, the most prominent respiratory ghosting
areas were not included in the ROI. The area of background was maintained
larger than 400 mm2. The lesion-to-liver signal difference-to-noise
ratio (SDNR) was calculated from the following equation:
where SIlesion is the SI of the lesion,
SIliver is the SI of the liver, and
SDnoise is the SD of the background noise. The
signal-to-noise ratio (SNR) was also calculated from the following equation:
To eliminate the bias from clustered data, the SDNRs of lesions of the same
pathology in each patient were averaged.
Two experienced gastrointestinal radiologists who had 5 and 2 years of
experience in MRI of the abdomen and were not involved in the initial
interpretation of the examinations jointly reviewed the four sets of MR images
in random order and drew consensus results for all the MRI studies. The
observers recorded the presence and anatomic segment of each lesion on
standardized data sheets to compare the results of the four image sets. The
observers recorded the conspicuity of the lesion in each image set based on a
4-point rating scale. A score of 1 indicated that the lesion was invisible; 2,
poorly visible; 3, fairly visible; and 4, clearly visible. All the images were
reviewed at a 1.5 x 1.5 K full PACS workstation (Centricity, GE
Healthcare).
Statistical Analysis
Paired-samples two-tailed Student's t tests were used to compare
the SDNR of every pair of each SPIO-enhanced GRE sequence. For statistical
analysis, all lesions were classified into either of the two groups according
to the potential capability of SPIO uptake by the lesions. HCCs, hemangiomas,
and focal eosinophilic infiltration were classified as SPIO-uptake lesions,
whereas all other lesions, including metastases and cysts, were classified as
non-SPIO-uptake lesions. In each group, a paired-samples Student's t
test was again performed to compare the SDNR of every pair of each
SPIO-enhanced GRE sequence. An independent sample Student's t test
was also performed to compare the SDNR between the SPIO-uptake and
non-SPIO-uptake lesions. The McNemar test was used to compare the sensitivity
between each pair of GRE sequences for the detection of focal hepatic lesions.
The Wilcoxon's signed rank test was used to compare the subjective lesion
conspicuity for depicted lesions. If a lesion was not seen on one sequence,
the conspicuity was recorded as 1, retrospectively, in that sequence. When a
lesion was not seen on all sequences, the conspicuity was regarded as 1 on all
sequences, retrospectively.
The SNRs of each group were compared by a paired-samples Student's
t test. Because we used a fixed dose of 1.4 mL of contrast material
for all patients, the total dose was inversely correlated with the total body
weight of the patients. To determine the effect of the variation of total
dose, patients' body weights were compared with liver SNR and lesion SDNR
using the Pearson's correlation coefficient. For all tests, a p value
of < 0.05 was considered to indicate a statistically significant
difference. Statistical analyses were performed with version 11 of SPSS
software (Statistical Package for the Social Sciences) for Microsoft
Windows.
Results
The mean SDNR was highest at a TE of 13.5 milliseconds (mean ± SD =
60 ± 24 milliseconds) and was significantly higher (p <
0.001) than those at TEs of 9 (55 ± 23), 18 (55 ± 22), and 22.5
(47 ± 19) milliseconds (Table
1). The SDNRs at TEs of 9 and 18 milliseconds were comparable
(p = 0.938) and were significantly higher than that at a TE of 22.5
milliseconds (p = 0.001). When the lesions were grouped into
SPIO-uptake or non-SPIO-uptake lesions, the SDNRs of the non-SPIO-uptake
lesions were higher than those of the SPIO-uptake lesions at TEs of 22.5
(p = 0.007) and 18 (p = 0.092) milliseconds. However, the
SDNRs were comparable between the non-SPIO-uptake and SPIO-uptake lesions at
TEs of 13.5 (p = 0.880) and 9 (p = 0.781) milliseconds.
View this table:
[in this window]
[in a new window]
|
TABLE 1: Mean Signal Difference-to-Noise Ratio on Gradient-Recalled Echo MRI for
Lesions With and Without Uptake of Superparamagnetic Iron Oxide (SPIO) at
Different TE Values
|
|
The mean SDNR of cholangiocarcinomas was highest at a TE of 18 milliseconds
and that of cysts was highest at a TE of 22.5 milliseconds
(Fig. 1). The mean SDNRs of
epithelioid hemangioendotheliomas, focal eosinophilic infiltrations, HCCs,
hemangiomas, and metastases were highest at a TE of 13.5 milliseconds. For
cysts, SDNRs at TEs of 18 (p = 0.009) and 22.5 (p = 0.014)
milliseconds were significantly higher than that at a TE of 9 milliseconds.
For HCCs, the SDNR at a TE of 13.5 milliseconds was significantly higher than
those at TEs of 9 (p = 0.007), 18 (p < 0.001), and 22.5
(p < 0.001) milliseconds, and the SDNR on images obtained using a
TE of 22.5 milliseconds was significantly lower than those obtained using TEs
of 9 (p < 0.001), 13.5 (p < 0.001), and 18 (p
< 0.001) milliseconds. For hemangiomas, the SDNR at a TE of 13.5
milliseconds was significantly higher than those at TEs of 18 (p =
0.002) and 22.5 (p < 0.001) milliseconds. The SDNR at a TE of 22.5
milliseconds was significantly lower than those at TEs of 9 (p =
0.018), 13.5 (p < 0.001), and 18 (p < 0.001)
milliseconds. For metastases, the SDNR at a TE of 22.5 milliseconds was
significantly lower than those at TEs of 9 (p = 0.002), 13.5
(p = 0.001), and 18 (p = 0.017) milliseconds, and there was
no significant difference in the SDNR for images obtained using TEs of 9,
13.5, and 22.5 (p > 0.3) milliseconds.
View larger version (34K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1 —Bar graph shows mean signal difference-to-noise ratios (SDNR)
of lesions on gradient-recalled MRI by lesion type. Mean SDNR of
cholangiocarcinomas (CCC) was highest at TE of 18 milliseconds and that of
cysts was highest at TE of 22.5 milliseconds. Mean SDNR of epithelioid
hemangioendotheliomas (EHE), focal eosinophilic infiltrations (FEI),
hepatocellular carcinomas (HCC), hemangiomas (HMG), and metastases (Mets) was
highest at TE of 13.5 milliseconds.
|
|
For SPIO-uptake lesions, the SDNR at a TE of 13.5 milliseconds was
significantly higher than those at TEs of 9 (p = 0.003), 18
(p < 0.001), and 22.5 (p < 0.001) milliseconds
(Table 2 and Figs.
2A,
2B,
2C,
2D,
3A,
3B,
3C, and
3D). The SDNRs of SPIO-uptake
lesions at TEs of 9 and 18 milliseconds were comparable (p = 0.196)
and higher than that at a TE of 22.5 milliseconds (p < 0.001). For
non-SPIO-uptake lesions, the SDNR was highest at a TE of 18 milliseconds but
was comparable with that at a TE of 13.5 milliseconds (p = 0.647),
and both were significantly higher than that at a TE of 22.5 milliseconds
(p = 0.007 and 0.015, respectively). The SDNR at a TE of 9
milliseconds was comparable with those at TEs of 13.5, 18, and 22.5
milliseconds (Figs. 4A,
4B,
4C, and
4D).
View this table:
[in this window]
[in a new window]
|
TABLE 2: p Values for Comparison of the Signal Difference-to-Noise Ratio
(SDNR) Between Each TE Group in Lesions With and Without Uptake of
Superparamagnetic Iron Oxide (SPIO) on Gradient-Recalled Echo MRI
|
|
View larger version (181K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —68-year-old man with hepatocellular carcinoma (HCC) and
cirrhosis. Superparamagnetic iron oxide (SPIO)-enhanced gradient-recalled echo
images obtained with TEs of 9 (A), 13.5 (B), 18 (C), and
22.5 (D) milliseconds. Signal intensities of both HCC (arrow)
and liver gradually decreased as TE was lengthened, but signal decrease of
liver was more prominent between images obtained with TEs of 9 and 13.5
milliseconds. Signal difference-to-noise ratio was highest at TE of 13.5
milliseconds.
|
|
View larger version (177K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —68-year-old man with hepatocellular carcinoma (HCC) and
cirrhosis. Superparamagnetic iron oxide (SPIO)-enhanced gradient-recalled echo
images obtained with TEs of 9 (A), 13.5 (B), 18 (C), and
22.5 (D) milliseconds. Signal intensities of both HCC (arrow)
and liver gradually decreased as TE was lengthened, but signal decrease of
liver was more prominent between images obtained with TEs of 9 and 13.5
milliseconds. Signal difference-to-noise ratio was highest at TE of 13.5
milliseconds.
|
|
View larger version (165K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C —68-year-old man with hepatocellular carcinoma (HCC) and
cirrhosis. Superparamagnetic iron oxide (SPIO)-enhanced gradient-recalled echo
images obtained with TEs of 9 (A), 13.5 (B), 18 (C), and
22.5 (D) milliseconds. Signal intensities of both HCC (arrow)
and liver gradually decreased as TE was lengthened, but signal decrease of
liver was more prominent between images obtained with TEs of 9 and 13.5
milliseconds. Signal difference-to-noise ratio was highest at TE of 13.5
milliseconds.
|
|
View larger version (155K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2D —68-year-old man with hepatocellular carcinoma (HCC) and
cirrhosis. Superparamagnetic iron oxide (SPIO)-enhanced gradient-recalled echo
images obtained with TEs of 9 (A), 13.5 (B), 18 (C), and
22.5 (D) milliseconds. Signal intensities of both HCC (arrow)
and liver gradually decreased as TE was lengthened, but signal decrease of
liver was more prominent between images obtained with TEs of 9 and 13.5
milliseconds. Signal difference-to-noise ratio was highest at TE of 13.5
milliseconds.
|
|
View larger version (155K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —40-year-old woman with hemangioma. Superparamagnetic iron
oxide (SPIO)-enhanced gradient-recalled echo images obtained with TEs of 9
(A), 13.5 (B), 18 (C), and 22.5 (D) milliseconds.
Signal intensity (SI) of both hemangioma (arrow) and liver decreased
as TE was lengthened, but loss of SI of lesion is more prominent because
contrast agent is pooling in lesion; hence, lesion conspicuity is decreased on
images obtained with longer TEs.
|
|
View larger version (155K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B —40-year-old woman with hemangioma. Superparamagnetic iron
oxide (SPIO)-enhanced gradient-recalled echo images obtained with TEs of 9
(A), 13.5 (B), 18 (C), and 22.5 (D) milliseconds.
Signal intensity (SI) of both hemangioma (arrow) and liver decreased
as TE was lengthened, but loss of SI of lesion is more prominent because
contrast agent is pooling in lesion; hence, lesion conspicuity is decreased on
images obtained with longer TEs.
|
|
View larger version (140K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3C —40-year-old woman with hemangioma. Superparamagnetic iron
oxide (SPIO)-enhanced gradient-recalled echo images obtained with TEs of 9
(A), 13.5 (B), 18 (C), and 22.5 (D) milliseconds.
Signal intensity (SI) of both hemangioma (arrow) and liver decreased
as TE was lengthened, but loss of SI of lesion is more prominent because
contrast agent is pooling in lesion; hence, lesion conspicuity is decreased on
images obtained with longer TEs.
|
|
View larger version (115K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3D —40-year-old woman with hemangioma. Superparamagnetic iron
oxide (SPIO)-enhanced gradient-recalled echo images obtained with TEs of 9
(A), 13.5 (B), 18 (C), and 22.5 (D) milliseconds.
Signal intensity (SI) of both hemangioma (arrow) and liver decreased
as TE was lengthened, but loss of SI of lesion is more prominent because
contrast agent is pooling in lesion; hence, lesion conspicuity is decreased on
images obtained with longer TEs.
|
|
View larger version (172K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A —51-year-old man with metastasis from rectal carcinoma.
Superparamagnetic iron oxide (SPIO)-enhanced gradient-recalled echo images
obtained with TEs of 9 (A), 13.5 (B), 18 (C), and 22.5
(D) milliseconds. There is no remarkable signal loss in small
metastasis (arrow) on images obtained with longer TEs, and lesion
conspicuity is comparable in all images; however, lesion looks smaller as TEs
are increased.
|
|
View larger version (163K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B —51-year-old man with metastasis from rectal carcinoma.
Superparamagnetic iron oxide (SPIO)-enhanced gradient-recalled echo images
obtained with TEs of 9 (A), 13.5 (B), 18 (C), and 22.5
(D) milliseconds. There is no remarkable signal loss in small
metastasis (arrow) on images obtained with longer TEs, and lesion
conspicuity is comparable in all images; however, lesion looks smaller as TEs
are increased.
|
|
View larger version (157K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4C —51-year-old man with metastasis from rectal carcinoma.
Superparamagnetic iron oxide (SPIO)-enhanced gradient-recalled echo images
obtained with TEs of 9 (A), 13.5 (B), 18 (C), and 22.5
(D) milliseconds. There is no remarkable signal loss in small
metastasis (arrow) on images obtained with longer TEs, and lesion
conspicuity is comparable in all images; however, lesion looks smaller as TEs
are increased.
|
|
View larger version (147K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4D —51-year-old man with metastasis from rectal carcinoma.
Superparamagnetic iron oxide (SPIO)-enhanced gradient-recalled echo images
obtained with TEs of 9 (A), 13.5 (B), 18 (C), and 22.5
(D) milliseconds. There is no remarkable signal loss in small
metastasis (arrow) on images obtained with longer TEs, and lesion
conspicuity is comparable in all images; however, lesion looks smaller as TEs
are increased.
|
|
In terms of detection sensitivity, 124 lesions (95%) were detected on
images obtained at a TE of 9, 125 (95%) at a TE of 13.5, 122 (93%) at a TE of
18, and 120 (92%) at a TE of 22.5 milliseconds. The difference did not reach
statistical significance in any comparison. Among the 11 lesions that were not
detected on images at a TE of 22.5 milliseconds, 1 HCC, 1 metastasis, and 2
hemangiomas were seen on images at TEs of 18, 13.5, and 9 milliseconds. One
HCC was detected at TEs of 18 and 13.5 milliseconds (Figs.
5A,
5B,
5C, and
5D). One metastasis and two
hemangiomas were not detected on images obtained using a TE of 18
milliseconds, but these lesions were detected on the other images. Of the
lesions that were not detected on the images obtained at a TE of 13
milliseconds, none was detected on the other images. One small metastasis that
was not detected on images at a TE of 9 milliseconds was detected on the other
GRE images, but this lesion was seen retrospectively. One well-differentiated
HCC of 1.5 cm was not seen on any GRE image, but it was seen on an unenhanced
T2-weighted fast spin-echo image (Figs.
6A,
6B,
6C,
6D, and
6E). An HCC of 1.7 cm and a
hemangioma of 1.5 cm were not detected on GRE images obtained using a TE of
22.5 milliseconds. All other lesions not detected on any images were 1 cm or
smaller in diameter.
View larger version (135K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A —46-year-old man with hepatocellular carcinoma (HCC).
Superparamagnetic iron oxide (SPIO)-enhanced steady-state gradient-recalled
echo images obtained with TEs of 9 (A), 13.5 (B), 18 (C),
and 22.5 (D) milliseconds. Signal intensity of small HCC is markedly
decreased as TE increases due to contrast uptake within lesion and prominent
susceptibility effect; hence, lesion is poorly shown on images obtained with
longer TE.
|
|
View larger version (112K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B —46-year-old man with hepatocellular carcinoma (HCC).
Superparamagnetic iron oxide (SPIO)-enhanced steady-state gradient-recalled
echo images obtained with TEs of 9 (A), 13.5 (B), 18 (C),
and 22.5 (D) milliseconds. Signal intensity of small HCC is markedly
decreased as TE increases due to contrast uptake within lesion and prominent
susceptibility effect; hence, lesion is poorly shown on images obtained with
longer TE.
|
|
View larger version (105K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5C —46-year-old man with hepatocellular carcinoma (HCC).
Superparamagnetic iron oxide (SPIO)-enhanced steady-state gradient-recalled
echo images obtained with TEs of 9 (A), 13.5 (B), 18 (C),
and 22.5 (D) milliseconds. Signal intensity of small HCC is markedly
decreased as TE increases due to contrast uptake within lesion and prominent
susceptibility effect; hence, lesion is poorly shown on images obtained with
longer TE.
|
|
View larger version (129K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5D —46-year-old man with hepatocellular carcinoma (HCC).
Superparamagnetic iron oxide (SPIO)-enhanced steady-state gradient-recalled
echo images obtained with TEs of 9 (A), 13.5 (B), 18 (C),
and 22.5 (D) milliseconds. Signal intensity of small HCC is markedly
decreased as TE increases due to contrast uptake within lesion and prominent
susceptibility effect; hence, lesion is poorly shown on images obtained with
longer TE.
|
|
View larger version (154K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6A —68-year-old man with two hepatocellular carcinomas (HCCs)
(Edmondson grade 2) and cirrhosis. Superparamagnetic iron oxide
(SPIO)-enhanced gradient-recalled echo images obtained with TEs of 9
(A), 13.5 (B), 18 (C), and 22.5 (D) milliseconds
and unenhanced T2-weighted fast spin-echo image (TR/TE, 6,000/105) (E).
Nodular HCC at lateral segment of liver (arrows) shows gradual signal
decrease as TE increases. Another small HCC (Edmondson grade 1-2) at medial
segment (arrowhead, E) was depicted on unenhanced T2-weighted
fast spin-echo image only.
|
|
View larger version (166K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6B —68-year-old man with two hepatocellular carcinomas (HCCs)
(Edmondson grade 2) and cirrhosis. Superparamagnetic iron oxide
(SPIO)-enhanced gradient-recalled echo images obtained with TEs of 9
(A), 13.5 (B), 18 (C), and 22.5 (D) milliseconds
and unenhanced T2-weighted fast spin-echo image (TR/TE, 6,000/105) (E).
Nodular HCC at lateral segment of liver (arrows) shows gradual signal
decrease as TE increases. Another small HCC (Edmondson grade 1-2) at medial
segment (arrowhead, E) was depicted on unenhanced T2-weighted
fast spin-echo image only.
|
|
View larger version (154K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6C —68-year-old man with two hepatocellular carcinomas (HCCs)
(Edmondson grade 2) and cirrhosis. Superparamagnetic iron oxide
(SPIO)-enhanced gradient-recalled echo images obtained with TEs of 9
(A), 13.5 (B), 18 (C), and 22.5 (D) milliseconds
and unenhanced T2-weighted fast spin-echo image (TR/TE, 6,000/105) (E).
Nodular HCC at lateral segment of liver (arrows) shows gradual signal
decrease as TE increases. Another small HCC (Edmondson grade 1-2) at medial
segment (arrowhead, E) was depicted on unenhanced T2-weighted
fast spin-echo image only.
|
|
View larger version (146K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6D —68-year-old man with two hepatocellular carcinomas (HCCs)
(Edmondson grade 2) and cirrhosis. Superparamagnetic iron oxide
(SPIO)-enhanced gradient-recalled echo images obtained with TEs of 9
(A), 13.5 (B), 18 (C), and 22.5 (D) milliseconds
and unenhanced T2-weighted fast spin-echo image (TR/TE, 6,000/105) (E).
Nodular HCC at lateral segment of liver (arrows) shows gradual signal
decrease as TE increases. Another small HCC (Edmondson grade 1-2) at medial
segment (arrowhead, E) was depicted on unenhanced T2-weighted
fast spin-echo image only.
|
|
View larger version (130K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6E —68-year-old man with two hepatocellular carcinomas (HCCs)
(Edmondson grade 2) and cirrhosis. Superparamagnetic iron oxide
(SPIO)-enhanced gradient-recalled echo images obtained with TEs of 9
(A), 13.5 (B), 18 (C), and 22.5 (D) milliseconds
and unenhanced T2-weighted fast spin-echo image (TR/TE, 6,000/105) (E).
Nodular HCC at lateral segment of liver (arrows) shows gradual signal
decrease as TE increases. Another small HCC (Edmondson grade 1-2) at medial
segment (arrowhead, E) was depicted on unenhanced T2-weighted
fast spin-echo image only.
|
|
The overall subjective ratings of lesion conspicuity were comparable among
images obtained using TEs of 9, 13.5, and 18 milliseconds and were
significantly better than those of images obtained using a TE of 22.5
milliseconds (p = 0.005-< 0.001)
(Table 3). For the SPIO-uptake
lesions, the subjective rating of lesion conspicuity was best on the images
obtained with a TE of 13.5 milliseconds, which was comparable with that of the
images obtained with a TE of 9 milliseconds, followed by that obtained with a
TE of 18 milliseconds (p < 0.02), and then lowest on the images
obtained with a TE of 22.5 milliseconds (p < 0.001). For the
non-SPIO-uptake lesions, the subjective ratings of the lesion conspicuity were
not significantly different in any comparison.
The mean SNR of the liver was highest at a TE of 9 milliseconds (14.7
± 6.5) and was significantly higher than those at TEs of 13.5 (8.6
± 3.2, p < 0.001), 18 (5.9 ± 1.8, p <
0.001), and 22.5 (4.9 ± 1.3, p < 0.001) milliseconds. The
mean SNR at a TE of 13.5 milliseconds was significantly higher (p
<0.001) than those at TEs of 18 and 22.5 milliseconds. The mean SNR at a TE
of 18 milliseconds was significantly higher (p < 0.001) than that
at a TE of 22.5 milliseconds. There was no significant correlation between the
liver SNR and patient body weight or total dose of ferucarbotran in all GRE
sequences (R = -0.105 to -0.021, p >0.05).
Discussion
The results of this study showed that the SDNR was significantly higher
overall at a TE of 13.5 milliseconds for lesions. In addition, the subjective
lesion conspicuity was comparable between TEs of 9 and 13.5 milliseconds, and
the overall lesion sensitivity was not significantly different among the four
image sets. However, the SDNR and lesion conspicuity were significantly lower
and the largest number of lesions were not detected on images obtained with a
TE of 22.5 milliseconds. Most lesions looked smaller as the TE increased and
the magnetic susceptibility effect became more prominent, especially at the
dome of the liver and the margin of the liver abutting the gastric or colonic
air. For the lesions that were not seen only at a TE of 22.5 milliseconds, the
detection failure was attributed to the prominent signal decrease and
susceptibility effect, whereas failure in the detection of the remaining
lesions was attributed to the inherent poor contrast of very small lesions and
the substantial SPIO uptake in some of the lesions, including HCCs,
hemangiomas, and focal eosinophilic infiltrations. The numbers of acquisitions
for imaging the entire liver needed to be increased as TE was
increased—from 2 acquisitions at a TE of 9 milliseconds to 5
acquisitions at a TE of 22.5 milliseconds. Therefore, we suggest that
SPIO-enhanced GRE images should not be obtained at a TE of 22.5 milliseconds
or longer for the detection of focal SPIO lesions.
Our results also showed that the SDNR of the focal liver lesions varied
according to the lesion's capability for SPIO uptake. Although the lesions
that may take up SPIO agents (e.g., hemangiomas and HCCs) showed the highest
SDNR at a TE of 13.5 milliseconds, the lesions that may not take up SPIO
agents (metastases and cysts) showed a comparably higher SDNR at TEs of 18 and
13.5 milliseconds. This was because the signal decrease was more pronounced in
SPIO-uptake lesions than in non-SPIO-uptake lesions when the TE was increased
to 18 milliseconds or more. Because hemangiomas and HCCs showed a more
prominent signal drop on the images obtained using a very long TE (18 and 22.5
milliseconds), comparison of SPIO-enhanced GRE images obtained using a very
long TE and those obtained using a moderately long TE (9 and 13.5
milliseconds) enables more accurate characterization of the lesions.
It is well known that benign hepatic lesions, including hemangiomas, focal
nodular hyperplasia (FNH), and hepatic adenomas, can take up SPIO agents
[15-19].
Previous studies have shown that SPIO-enhanced MRI reflects Kupffer cell count
in HCCs and dysplastic nodules and is useful for estimation of histologic
grading in HCC; and the conspicuity of well-differentiated HCC or dysplastic
nodules is decreased on the SPIO-enhanced fast spin-echo or GRE images
obtained using a TE of 8.4-10 milliseconds
[20,
21]. In our study, one
well-differentiated HCC that was well depicted on an unenhanced T2-weighted
fast spin-echo image was not depicted on the contrast-enhanced GRE images
because of SPIO uptake. However, even the HCC lesions that showed
hyperintensity on the GRE images at TEs of 9 or 13.5 milliseconds showed a
substantial decrease in SI on images obtained using a longer TE (18 and 22.5
milliseconds) in contrast with the non-HCC lesions such as metastases. The
possibility of minimally retained Kupffer cell activity in the HCC or the
increased vascularity of the tumor might have contributed to this finding.
The results of the present study deserve comparison with the results of the
studies by Ward et al. [13]
and Matsuo et al. [12]. Matsuo
et al. compared GRE images at TEs of 1.4, 4.2, 6, 8, and 10 milliseconds for
the detection of HCCs and metastases using ferumoxides at a dose of 10 µmol
Fe/kg. In their study, GRE sequences at a TE of 8 milliseconds showed the
highest sensitivity and liver-to-lesion contrast, but the difference was not
significant when compared with GRE images obtained at a TE of 10 milliseconds.
The flip angle used in their study was 70-90° and might have a higher T1
effect than the flip angle of 30° used in our study. In the study by Ward
et al. [13], using ferumoxides
at a dose of 7.5 µmol Fe/kg, SPIO-enhanced GRE images that were obtained at
a TE of 15 milliseconds and at the flip angle of 30° showed significantly
higher liver-to-lesion contrast than GRE images at a TE of 11 milliseconds,
but the sensitivity was not significantly different for the detection of
colorectal cancer metastases. These results are comparable to those of our
study.
In the present study, we compared TEs of 9, 13.5, 18, and 22.5 milliseconds
to encompass the TE range used in previous studies but not directly compared
and to obtain in-phase images to reduce the ring cancellation artifacts from
chemical shift effects along the liver surface
[4-8,
10-14].
We used a flip angle of 30° to minimize the T1 effect and reduce pulsation
artifact and to compare our data with the results of a previous study
[13]. We also used a fixed
bandwidth for all TE values. If we reduced the bandwidth as we increased TE,
the SNR could be improved for images obtained with a long TE. However, if we
did so, the disadvantages of increased artifact at the periphery of the liver
and decreased slices per breath-hold might be more severe. We speculate that
using a higher frequency encoding (e.g., 512 instead of 256) might reduce the
apparent susceptibility at the periphery of the liver and improve the
differentiation of a small lesion with a branch of a vessel by improving the
spatial resolution. However, we maintained a constant frequency encoding in
this study to directly compare the TE effect.
We used 1.4 mL (7.0-14.5 µmol Fe/kg) of ferucarbotran. A prior study
showed that ferucarbotran, administered in a dose of 7.0-12.9 µmol of iron
per kilogram of body weight, had similar T1 and T2 effects on the liver when
compared with ferumoxides administered in a dose of 15 µmol/kg
[22]. In this context, the
dose used in our study might be somewhat higher than those used by Matsuo et
al. [12] and Ward et al.
[13]. Nonetheless, our results
showed that lesion contrast can be increased until a TE of 18 milliseconds for
metastases and until a TE of 13.5 milliseconds for HCCs and hemangiomas. This
is also comparable with the findings of another study using a low dose (7.5
µmol Fe/kg) of ferumoxides and a 1.0-T system: Liver-to-lesion
contrast-to-noise ratio was increased until a TE of 26 milliseconds
[23].
Our results also showed that the patients' body weights were not correlated
with the liver SNR or lesion SDNR on any GRE sequence. This suggests that the
variation of dose range used in this study does not affect the liver SNR or
lesion SDNR. These results are also similar to those of previous reports in
which the efficacy of ferucarbotran was comparable for the doses of 8 and 16
µmol Fe/kg [24,
25].
A limitation of our study is that many lesions were not pathologically
proven. Typical clinical and laboratory findings in combination with disease
progression, as depicted on follow-up images, were used as the diagnostic
criteria in these patients. Although we applied these diagnostic criteria as
strictly as possible to exclude false-positive lesions, we might have missed
small lesions. The relative high sensitivity of each GRE image set might be
attributed to this lack of histologic confirmation. However, because we
compared the SDNR and detection rate among the four image sets obtained under
the same conditions only for the confirmed lesions, we believe that this
limitation would not alter the validity of this study.
A second limitation is that we did not analyze observer variability but,
rather, used consensus interpretation for lesion detection. A multiobserver
analysis would have estimated better the sensitivity of each technique for
depicting lesions and would have more accurately predicted differences between
them in clinical practice. However, our consensus interpretation is still
valid for evaluating the choice of optimal TE for SPIO-enhanced imaging.
A third limitation is that we did not randomize the order in which we
obtained GRE images with a different TE; instead, we acquired the images with
a TE of 22.5 milliseconds first and those with a TE of 9 milliseconds last. We
chose this consistent order to avoid ambiguity between potential benefits to
image contrast from more complete vascular clearance of contrast agent versus
increased susceptibility artifacts from a longer TE; the signal drop was most
profound on the images obtained with a TE of 22.5 milliseconds, even though
vascular clearance might have been less. In addition, the time difference of
20 minutes between the first and the last image set probably had little effect
on tissue contrast, particularly considering that a prior study using the same
SPIO agent found no apparent differences between 10- and 40-minute
contrast-enhanced MR images for evaluating focal lesions
[25].
Because of the small number of histologically proven HCCs, we could not
determine whether the decreased SI and SDNR of HCCs at very long TEs of 18 and
22.5 milliseconds might be related to the histologic grade or other
microscopic characteristics. In many reports, researchers have shown that
benign hepatocellular nodules, including FNH and hepatic adenomas, decrease in
SI on SPIO-enhanced MR images
[26-31].
Because none of these lesions were included in this study, we cannot state how
the signal characteristics of those lesions might be changed on images
obtained with the TE range used in our study.
We performed dynamic MRI as part of our post-ferucarbotran administration
protocol, but we did not analyze these images for this study. On the
T1-weighted SPIO-enhanced dynamic MR images, features that are useful for the
characterization of focal lesions include peripheral globular enhancement of
hemangiomas, rim enhancement of metastases, and early increased enhancement of
HCCs
[32-34].
We noted these findings in many lesions in our study, but this was not a
purpose of this study.
A faster imaging sequence, such as diffusion-weighted echo-planar imaging
[35], gradient and spin-echo
[36], or balanced steady-state
imaging (e.g., balanced fast-field echo, true fast imaging with steady-state
free precession [FISP], or fast imaging employing steady-state acquisition
[FIESTA]), might be useful as a faster method with fewer artifacts that has
susceptibility contrast, but we have no data to address the potential utility
of these techniques because they were not available at the time of this
study.
In summary, our results showed that a TE of 13.5 milliseconds provided the
highest SDNR for SPIO-enhanced GRE images, but the SDNR of liver lesions can
vary according to the lesion's potential capacity for SPIO uptake. The
subjective lesion conspicuity and sensitivity were comparable when the TE was
in the range of 9-18 milliseconds, but the SDNR of the lesions that may take
up SPIO agents substantially decreased on GRE images at TEs of 18 and 22.5
milliseconds.
In conclusion, we suggest that TE values can be chosen from 9 to 18
milliseconds for SPIO-enhanced MRI in consideration of scanning time and image
quality. Our results also suggest that comparing the GRE images with a
moderately long TE (9-13.5 milliseconds) and a very long TE (18-22.5
milliseconds) may be useful for characterization of focal hepatic lesions.
References
- Semelka RC, Helmberger TK. Contrast agents for MR imaging of the
liver. Radiology 2001;218
: 27-38[Abstract/Free Full Text]
- Reimer P, Tombach B. Hepatic MRI with SPIO: detection and
characterization of focal liver lesions. Eur Radiol1998; 8:1198
-1204[CrossRef][Medline]
- Reimer P, Balzer T. Ferucarbotran (Resovist): a new clinically
approved RES-specific contrast agent for contrast-enhanced MRI of the
liver—properties, clinical development, and applications. Eur
Radiol 2003; 13:1266
-1276[Medline]
- Fretz CJ, Elizondo G, Weissleder R, Hahn PF, Stark DD, Ferrucci JT
Jr. Superparamagnetic iron oxide-enhanced MR imaging: pulse sequence
optimization for detection of liver cancer. Radiology1989; 172:393
-397[Abstract/Free Full Text]
- Bellin MF, Zaim S, Auberton E, et al. Liver metastases: safety and
efficacy of detection with superparamagnetic iron oxide in MR imaging.
Radiology 1994;193
: 657-663[Abstract/Free Full Text]
- Oudkerk M, van den Heuvel AG, Wielopolski PA, Schmitz PI, Borel
Rinkes IH, Wiggers T. Hepatic lesions: detection with ferumoxide-enhanced
T1-weighted MR imaging. Radiology 1997;203
: 449-456[Abstract/Free Full Text]
- Ward J, Chen F, Guthrie JA, et al. Hepatic lesion detection after
superparamagnetic iron oxide enhancement: comparison of five T2-weighted
sequences at 1.0 T by using alternative-free response receiver operating
characteristic analysis. Radiology 2000;214
: 159-166[Abstract/Free Full Text]
- Arbab AS, Ichikawa T, Araki T, et al. Detection of hepatocellular
carcinoma and its metastases with various pulse sequences using
superparamagnetic iron oxide (SHU-555-A). Abdom
Imaging 2000; 25:151
-158[CrossRef][Medline]
- Sugihara S, Suto Y, Kamba M, Ogawa T. Comparison of various
techniques of iron oxide-enhanced breath-hold MR imaging of hepatocellular
carcinoma. Clin Imaging 2001;25
: 104-109[CrossRef][Medline]
- Kanematsu M, Itoh K, Matsuo M, et al. Malignant hepatic tumor
detection with ferumoxides-enhanced MR imaging with a 1.5-T system: comparison
of four imaging pulse sequences. J Magn Reson Imaging2001; 13:249
-257[CrossRef][Medline]
- Kim SH, Choi D, Lim JH, et al. Optimal pulse sequence for
ferumoxides-enhanced MR imaging used in the detection of hepatocellular
carcinoma: a comparative study using seven pulse sequences. Korean
J Radiol 2002; 3:87
-97[Medline]
- Matsuo M, Kanematsu M, Itoh K, et al. Detection of malignant
hepatic tumors with ferumoxides-enhanced MRI: comparison of five
gradient-recalled echo sequences with different TEs.
AJR 2004; 182:235
-242[Abstract/Free Full Text]
- Ward J, Guthrie JA, Wilson D, et al. Colorectal hepatic metastases:
detection with SPIO-enhanced breath-hold MR imaging—comparison of
optimized sequences. Radiology 2003;228
: 709-718[Abstract/Free Full Text]
- Van Beers BE, Lacrosse M, Jamart J, et al. Detection and segmental
location of malignant hepatic tumors: comparison of ferumoxides-enhanced
gradient-echo and T2-weighted spin-echo MR imaging.
AJR 1997; 168:713
-717[Abstract/Free Full Text]
- van Gansbeke D, Metens TM, Matos C, et al. Effects of AMI-25 on
liver vessels and tumors on T1-weighted turbo-field-echo images: implications
for tumor characterization. J Magn Reson Imaging1997; 7:482
-489[Medline]
- Grangier C, Tourniaire J, Mentha G, et al. Enhancement of liver
hemangiomas on T1-weighted MR SE images by superparamagnetic iron oxide
particles. J Comput Assist Tomogr 1994;18
: 888-896[Medline]
- Nakayama M, Yamashita Y, Mitsuzaki K, et al. Improved tissue
characterization of focal liver lesions with ferumoxide-enhanced T1 and
T2-weighted MR imaging. J Magn Reson Imaging2000; 11:647
-654[CrossRef][Medline]
- Paley MR, Mergo PJ, Torres GM, Ros PR. Characterization of focal
hepatic lesions with ferumoxides-enhanced T2-weighted MR imaging.
AJR 2000; 175:159
-163[Abstract/Free Full Text]
- Kim JH, Kim MJ, Suh SH, Chung JJ, Yoo HS, Lee JT. Characterization
of focal hepatic lesions with ferumoxides-enhanced MR imaging: utility of
T1-weighted spoiled gradient recalled echo images using different echo times.
J Magn Reson Imaging 2002;15
: 573-583[CrossRef][Medline]
- Imai Y, Murakami T, Yoshida S, et al. Superparamagnetic iron
oxide-enhanced magnetic resonance images of hepatocellular carcinoma:
correlation with histological grading. Hepatology2000; 32:205
-212[CrossRef][Medline]
- Lim JH, Choi D, Cho SK, et al. Conspicuity of hepatocellular
nodular lesions in cirrhotic livers at ferumoxides-enhanced MR imaging:
importance of Kupffer cell number. Radiology2001; 220:669
-676[Abstract/Free Full Text]
- Chen F, Ward J, Robinson PJ. MR imaging of the liver and spleen: a
comparison of the effects on signal intensity of two superparamagnetic iron
oxide agents. Magn Reson Imaging 1999;17
: 549-556[CrossRef][Medline]
- Scott J, Ward J, Guthrie JA, Wilson D, Robinson PJ. MRI of liver: a
comparison of CNR enhancement using high dose and low dose ferumoxide infusion
in patients with colorectal liver metastases. Magn Reson
Imaging 2000; 18:297
-303[CrossRef][Medline]
- Kopp AF, Laniado M, Dammann F, et al. MR imaging of the liver with
Resovist: safety, efficacy, and pharmacodynamic properties.
Radiology 1997;204
: 749-756[Abstract/Free Full Text]
- Shamsi K, Balzer T, Saini S, et al. Superparamagnetic iron oxide
particles (SH U 555 A): evaluation of efficacy in three doses for hepatic MR
imaging. Radiology 1998;206
: 365-371[Abstract/Free Full Text]
- Grandin C, Van Beers BE, Robert A, Gigot JF, Geubel A, Pringot J.
Benign hepatocellular tumors: MRI after superparamagnetic iron oxide
administration. J Comput Assist Tomogr1995; 19:412
-418[Medline]
- Vogl TJ, Hammerstingl R, Schwarz W, et al. Superparamagnetic iron
oxide-enhanced versus gadolinium-enhanced MR imaging for differential
diagnosis of focal liver lesions. Radiology1996; 198:881
-887[Abstract/Free Full Text]
- Poeckler-Schoeniger C, Koepke J, Gueckel F, Sturm J, Georgi M. MRI
with superparamagnetic iron oxide: efficacy in the detection and
characterization of focal hepatic lesions. Magn Reson
Imaging 1999; 17:383
-392[CrossRef][Medline]
- Precetti-Morel S, Bellin MF, Ghebontni L, et al. Focal nodular
hyperplasia of the liver on ferumoxides-enhanced MR imaging: features on
conventional spin-echo, fast spin-echo and gradient-echo pulse sequences.
Eur Radiol 1999;9
: 1535-1542[CrossRef][Medline]
- Takeshita K, Nagashima I, Frui S, et al. Effect of
superparamagnetic iron oxide-enhanced MRI of the liver with hepatocellular
carcinoma and hyperplastic nodule. J Comput Assist
Tomogr 2002; 26:451
-455[CrossRef][Medline]
- Kim MJ, Kim JH, Chung JJ, Park MS, Lim JS, Oh YT. Focal hepatic
lesions: detection and characterization with combination gadolinium- and
superparamagnetic iron oxide-enhanced MR imaging.
Radiology 2003;228
: 719-726[Abstract/Free Full Text]
- Reimer P, Jahnke N, Fiebich M, et al. Hepatic lesion detection and
characterization: value of nonenhanced MR imaging, superparamagnetic iron
oxide-enhanced MR imaging, and spiral CT-ROC analysis.
Radiology 2000;217
: 152-158[Abstract/Free Full Text]
- Reimer P, Muller M, Marx C, et al. T1 effects of a bolus-injectable
superparamagnetic iron oxide, SH U 555 A: dependence on field strength and
plasma concentration—preliminary clinical experience with dynamic
T1-weighted MR imaging. Radiology 1998;209
: 831-836[Abstract/Free Full Text]
- Vogl TJ, Hammerstingl R, Schwarz W, et al. Magnetic resonance
imaging of focal liver lesions: comparison of the superparamagnetic iron oxide
Resovist versus gadolinium-DTPA in the same patient. Invest
Radiol 1996; 31:696
-708[CrossRef][Medline]
- Naganawa S, Sato C, Nakamura T, et al. Diffusion-weighted images of
the liver: comparison of tumor detection before and after contrast enhancement
with superparamagnetic iron oxide. J Magn Reson
Imaging 2005; 21:836
-840[CrossRef][Medline]
- Karantanas AH, Papanikolaou N. T2-weighted magnetic resonance
imaging of the liver: comparison of fat-suppressed GRASE with conventional
spin echo, fat-suppressed turbo spin echo, and gradient echo at 1.0 T.
Abdom Imaging 2001;26
: 139-145[CrossRef][Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
