AJR 2005; 184:842-846
© American Roentgen Ray Society
Utility of Breath-Hold Fast-Recovery Fast Spin-Echo T2 Versus Respiratory-Triggered Fast Spin-Echo T2 in Clinical Hepatic Imaging
Jimmy Huang1,
Steven S. Raman1,
Ngan Vuong1,
James W. Sayre1 and
David S. K. Lu1
1 All authors: Department of Radiology, David Geffen School of Medicine, Center
for the Health Sciences, UCLA Medical Center, BL-428 CHS, Box 951721, Los
Angeles, CA 90095-1721.
Received January 16, 2004;
accepted after revision July 26, 2004.
Address correspondence to S. S. Raman
(sraman{at}mednet.ucla.edu).
Abstract
OBJECTIVE. The objective of our study was to compare a breath-hold
fat-suppressed fast-recovery fast spin-echo (FSE) T2-weighted sequence with a
respiratory-triggered fat-suppressed FSE T2-weighted sequence to assess the
effect on image quality and lesion detection and characterization in clinical
hepatic imaging.
MATERIALS AND METHODS. Both the breath-hold fat-suppressed
fast-recovery FSE and respiratory-triggered fat-suppressed FSE T2-weighted
sequences were acquired in 46 patients. Two radiologists, blinded to clinical
data, independently evaluated randomized images from both sequences.
Qualitatively, images were graded on a 5-point scale for five different
characteristics. The number and location of lesions were recorded. The
confidence of detection and the confidence of characterization (solid vs
nonsolid) were graded on a 5-point scale. A consensus review using radiology,
clinical, and pathology data served as the standard. Receiver operating
characteristic (ROC) curve analysis (area under the ROC curve
[Az]) was used to compare each reviewer's interpretation
against the consensus interpretation. Quantitative analysis was performed by
calculating the liver signal-to-noise ratio (SNR), liver-to-spleen
contrast-to-noise ratio (CNR), and lesion-to-liver CNR. Both one- and
two-tailed Student's t tests were used to check for significance.
RESULTS. Qualitatively, both reviewers graded the breath-hold
fat-suppressed fast-recovery FSE T2-weighted sequence better than the
respiratory-triggered fat-suppressed FSE T2-weighted sequence on all five
characteristics (p < 0.005). Of 78 lesions detected, 29 were
characterized as solid; 47, nonsolid; and two, indeterminate. On ROC analysis,
there were no significant differences between the breath-hold fat-suppressed
fast-recovery FSE and respiratory-triggered fat-suppressed FSE T2-weighted
sequences in lesion detection (Az reviewer 1, 0.77 and
0.83, respectively, [p = 0.12]; Az reviewer 2,
0.84 and 0.80, respectively [p = 0.12]) or in lesion characterization
(Az reviewer 1, 0.86 and 0.92, respectively [p =
0.33]; Az reviewer 2, 0.90 and 0.91, respectively
[p = 0.79]). Quantitatively, liver SNRs, spleen CNRs, and lesion CNRs
(solid and nonsolid lesions) were significantly better on the breath-hold
fat-suppressed fast-recovery FSE T2-weighted images than on the
respiratory-triggered fat-suppressed FSE T2-weighted images (p <
0.005).
CONCLUSION. Breath-hold fat-suppressed fast-recovery FSE T2-weighted
images were of better quality than respiratory-triggered fat-suppressed FSE
T2-weighted images, and lesion detection and characterization were
comparable.
Introduction
For hepatic MRI, high-quality T2-weighted images are important for the
detection and characterization of focal lesions. A desirable T2-weighted
sequence should be performed rapidly while maintaining the high level of
tissue contrast provided by conventional respiratory-triggered fat-suppressed
fast spin-echo (FSE) sequences and should have reduced image artifacts. In the
mid 1990s traditional T2-weighted spin-echo sequences were supplanted by
signal-averaged FSE or turbo spin-echo sequences
[1], yielding significant
reduction in acquisition time while maintaining a high degree of tissue
contrast. However, these sequences were susceptible to a variety of
significant artifacts, even with techniques such as respiratory triggering,
often limiting the diagnostic usefulness in many patients
[2–4].
Breath-hold multishot FSE sequences were also developed, although their use
was limited by marginal performance with respect to the signal-to-noise ratio
(SNR) and the contrast-to-noise ratio (CNR)
[5]. A modified FSE with
fast-recovery technique has shown promise in decreasing imaging time and
imaging artifacts and maintaining high tissue contrast in initial reports
[6–8].
The purpose of this study was to compare the performance of the breath-hold
fat-suppressed fast-recovery FSE sequence with the respiratory-triggered FSE
sequence in routine clinical practice.
Materials and Methods
Patient Selection
An institutional review board exemption was obtained for this study. In
this retrospective study of MRI examinations performed between October 2002
and January 2003, 48 MR studies that included both breath-hold fat-suppressed
fast-recovery FSE and respiratory-triggered fat-suppressed FSE T2-weighted
sequences were reviewed. There were 46 patients, 18 women and 28 men with a
mean age of 59.2 years (age range, 30–85 years).
MRI Technique
MRI was performed with a superconducting 1.5-T scanner (Signa Horizon LX
[version 9.1], GE Healthcare) with a phased-array coil as the receiver. All MR
images were obtained in the axial plane. The section thickness was 8 mm with a
2-mm interslice gap for all T2-weighted sequences.
The protocol consisted of a conventional respiratory-triggered
fat-suppressed FSE sequence (TR/effective TE, 10,109/84; echo-train length, 8;
receiver bandwidth, 16 kHz; matrix, 256 x 256; number of excitations, 2;
field of view, 34 x 26 cm) and a breath-hold fat-suppressed
fast-recovery FSE sequence (TR/TE, 2,000/93; echo-train length, 16; receiver
bandwidth, 10.4 kHz; matrix, 256 x 265; field of view, 34 x 26 cm;
number of excitations, 2; acquisition time, 20–24 sec). Before each
T2-weighted sequence, manual shimming was performed and frequency-selective
fat suppression was applied.
Subsequently, T1-weighted FSE and dynamic gadolinium-enhanced
gradient-recalled echo imaging were performed. Because we do not rely
primarily on the T2 sequences for lesion characterization, only a single TE is
used. We use a combination of T1-weighted sequences with and without
gadolinium in addition to the T2 sequence for lesion characterization. For the
MR examinations, automated shimming was always used before data
acquisition.
Qualitative Image Analysis
Image quality.—Two experienced fellowship-trained abdominal
radiologists independently reviewed all MR images from each sequence during
two sessions separated by at least 4 weeks. The studies were reviewed in
random order using a random-number generator. All patient identifiers and
information about clinical history were withheld during the review of the
images. The reviewers graded images from both T2-weighted sequences for the
presence of artifacts (respiratory ghosting, bowel peristalsis, vascular
pulsation) using a 5-point scale: 1, severe; 2, moderate; 3, mild; 4, minimal;
and 5, absent. The quality of fat suppression was also graded (1, severely
inhomogeneous; 2, moderately inhomogeneous; 3, mildly inhomogeneous; 4, fairly
homogeneous; and 5, homogeneous). Finally, the degree of image sharpness was
assessed: 1, severely unsharp; 2, moderately unsharp; 3, mildly unsharp; 4,
fairly sharp; and 5, sharp.
Lesion detection and characterization.—Both reviewers
independently evaluated the overall number of hepatic lesions per sequence on
a segment-by-segment basis. Eight anatomic hepatic segments were defined on
the basis of the numbering system of Couinaud
[9]. For each sequence,
reviewers recorded the segmental location and the size of each lesion and then
assigned one of the following five possible confidence levels: 1, definitely
absent; 2, probably absent; 3, possibly present; 4, probably present; and 5,
definitely present. Reviewers then characterized each detected lesion on a
5-point scale (1, definitely nonsolid; 2, probably nonsolid; 3, indeterminate;
4, probably solid; and 5, definitely solid). Nonsolid lesions primarily
include cysts and hemangiomas. Solid lesions are the remaining lesions that
require biopsy or extensive follow-up. Solid lesions include hepatocellular
carcinoma, hepatocellular adenoma, metastasis, and focal nodular
hyperplasia.
Lesion verification.—Lesion verification was performed
approximately 4 weeks after the interpretations to minimize bias among the
reviewers. Three authors determined the reference standard after consensus
review using all available imaging (MRI, CT, and sonography), clinical (liver
function tests, 
-fetoprotein levels), and pathology (tissue diagnosis)
data. Thirteen patients had tissue diagnosis by either percutaneous biopsy or
surgery. In 29 patients, the consensus review detected a total of 78 lesions.
Only the consensus data were used in the final analysis.
Quantitative Image Analysis
After extensive instructions and training, the two reviewers performed the
region-of-interest measurements. An average of the three measurements of the
liver, spleen, and background was obtained to minimize errors. Quantitative
analysis was performed on T2-weighted MR images obtained with the two pulse
sequences using operator-defined region-of-interest measurements of mean
signal intensity in the liver, background noise, and hepatic lesions when
present. The signal intensity in the liver and spleen was measured in areas
devoid of large vessels, prominent artifacts, and focal changes. For liver
lesions, the region of interest was drawn to encompass as much of the lesion
as possible. Only lesions with an area of 50 mm2 or more were
included for quantitative evaluation to exclude any inaccuracies in signal
intensity measurements that may have resulted from partial volume effect.
The SD of the background signal intensity (SDair) was
measured in the largest possible region of interest positioned in the
phase-encoding direction outside the abdominal wall to account for any
artifacts. Then data were calculated using the following equations:
where SI is signal intensity, SNR is signal-to-noise ratio,
and CNR is contrast-to-noise ratio.
Statistical Analysis
A receiver operating characteristic (ROC) curve analysis was used to
compare lesion detection confidence and characterization for each reviewer
against consensus [10]. Each
reviewer's performance in interpreting the findings on each imaging sequence
was estimated by calculating the area under the ROC curve
(Az). Both one- and two-tailed Student's t tests
were used to check for significance for any differences between individual
reviewers and the consensus interpretation with respect to both the
respiratory-triggered FSE and fast-recovery FSE sequences.
Results
Qualitative Analysis
Image quality.—Overall, image quality was better on the
breath-hold fat-suppressed fast-recovery FSE sequence for both reviewers
(Table 1 and Figs.
1A,
1B,
2A,
2B,
3A,
3B,
4A, and
4B). The fast-recovery FSE
images had fewer image artifacts from respiratory ghosting, bowel peristalsis,
and vascular pulsation than the respiratory-triggered fat-suppressed FSE
sequence (p < 0.001). The breath-hold fat-suppressed fast-recovery
sequence also had more homogeneous fat suppression (p < 0.001).
The overall image sharpness and quality of the breath-hold fat-suppressed
fast-recovery sequence were significantly superior to those of the
conventional respiratory-triggered fat-suppressed FSE sequence (p
< 0.001).
View larger version (152K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —Fat-suppressed fast spin-echo T2-weighted images in 61-year-old
woman with pseudolesion in medial segment. Respiratory-triggered MR image
(TR/TE, 10,909/84) shows lesion as high signal intensity.
|
|
View larger version (154K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B. —Fat-suppressed fast spin-echo T2-weighted images in 61-year-old
woman with pseudolesion in medial segment. Breath-hold fast-recovery MR image
(2,416/91) reveals lesion as part of low-signal-intensity portal branches.
|
|
View larger version (140K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B. —Fat-suppressed fast spin-echo T2-weighted images in 67-year-old man
with hepatitis. Breath-hold fast-recovery MR image (2,416/97) reveals T2
high-signal-intensity lesion in segment VIII.
|
|
View larger version (125K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A. —Fat-suppressed fast spin-echo T2-weighted images in 69-year-old man
with ascites and prior radiofrequency ablation of liver lesion. Both reviewers
identified lesion in respiratory-triggered (A) (TR/TE, 10,909/84) and
breath-hold fast-recovery (B) (2,250/91) MR images, but confidence
interval was higher for B. Lesion depiction and sharpness are best in
B.
|
|
View larger version (127K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B. —Fat-suppressed fast spin-echo T2-weighted images in 69-year-old man
with ascites and prior radiofrequency ablation of liver lesion. Both reviewers
identified lesion in respiratory-triggered (A) (TR/TE, 10,909/84) and
breath-hold fast-recovery (B) (2,250/91) MR images, but confidence
interval was higher for B. Lesion depiction and sharpness are best in
B.
|
|
View larger version (140K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A. —Fat-suppressed fast spin-echo T2-weighted images in 73-year-old man
with end-stage liver disease and ascites. Respiratory ghosting and vascular
pulsations grossly degrade quality of respiratory-triggered MR image (TR/TE,
10,909/84).
|
|
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B. —Fat-suppressed fast spin-echo T2-weighted images in 73-year-old man
with end-stage liver disease and ascites. Artifacts visible in A are
absent in this breath-hold fast-recovery MR image (2,416/97), which depicts
liver parenchyma and surface in detail.
|
|
Lesion detection and characterization.—Seventy-eight lesions
were detected by consensus review. Twenty-nine were characterized as solid, 47
as nonsolid, and two as indeterminate. Both reviewers detected similar numbers
of lesions on respiratory-triggered FSE and fast-recovery FSE sequences,
without statistical significance (Table
2). Both reviewers had slightly higher confidence levels of lesion
detectability with breath-hold fast-recovery sequence than with conventional
respiratory-triggered FSE sequence. However, this difference was not
statistically significant. Lesion characterization was also similar for both
reviewers between sequences (Table
3).
Quantitative Analysis
In comparison to the respiratory-triggered FSE sequence, the fast-recovery
FSE sequence had a nearly twofold increase in the liver SNRs (p <
0.001), liver-to-spleen CNRs (p < 0.005), and mean lesion-to-liver
CNRs (p < 0.001) (Table
4).
Discussion
In this study, we have shown that breath-hold fat-suppressed fast-recovery
FSE T2-weighted images compare favorably with the established
respiratory-triggered fat-suppressed FSE T2-weighted images. The fast-recovery
FSE sequence provided significantly better qualitative and quantitative image
quality and significantly decreased image artifacts. Unlike prior breath-hold
multishot sequences in which tissue contrast was suboptimal
[5], the fast-recovery FSE
sequence actually had a nearly twofold increase in overall SNR,
liver-to-spleen CNR, and lesion-to-liver CNR when compared with the
respiratory-triggered FSE. This was not possible with prior breath-hold
sequences and corroborates the results of two prior reports
[8]. Unlike sequences such as
single-shot FSE sequences in which intermediate T2 signal lesions may be
obscured, the fast-recovery FSE sequence was similar to conventional
respiratory-triggered FSE for both lesion detection and characterization. In
most of the patients who were able to breath-hold, the fast-recovery FSE
sequence provides these significant advantages without measurable
disadvantages. From the perspective of both patient and radiologist, the time
savings is substantial (30–40 sec for fast-recovery FSE vs 3–5 min
for respiratory-triggered FSE) and is therefore desirable. Conventional
respiratory-triggered FSE T2-weighted imaging with fat saturation may be
reserved for those patients unable to breath-hold.
There are certain limitations to our study. First, for lesion detection and
characterization, our gold standard was suboptimal. However, a more rigorous
standard such as biopsy or intraoperative sonography would be impossible in
most patients, especially those with benign lesions. Consensus review using
available imaging, clinical, and pathology data more closely reflects routine
clinical practice and provides a reasonable standard because the relative
proportion of detected lesions per sequence is most important. Also, despite
the effort to characterize the lesions as cystic versus solid, we did not
differentiate benign solid from malignant solid lesions because that was not
the focus of this study. Finally, some bias was unavoidable because the
sequences have distinct appearances and the reviewers therefore were unblinded
to some degree. However, recall bias was minimized by a 4-week time interval
between reviews of the different sequences. Our 8-mm images are relatively
thick given today's technology. These images were acquired with the earlier
protocols that were used during the infancy stages of sequence
development.
The fast-recovery FSE sequence allows image acquisition in less time than
the respiratory-triggered FSE sequence without compromising image quality.
High-resolution imaging is absolutely feasible with this sequence. The slice
thickness can be decreased, consequently improving resolution, without
significantly prolonging the scanning time. Also, with the recent introduction
of 3-T MR scanners, thin-slice high-resolution images can be obtained in an
even shorter time with the fast-recovery FSE sequence.
In summary, the breath-hold fat-suppressed fast-recovery FSE sequence is a
robust T2-weighted sequence that may replace the conventional
respiratory-triggered fat-suppressed FSE T2-weighted sequence for imaging the
liver in most patients.
References
- Catasca JV, Mirowitz SA. T2-weighted MR imaging of the abdomen:
fast spin-echo vs conventional spin-echo sequences.
AJR 1994;162:61
-67[Abstract/Free Full Text]
- Rydberg JN, Lomas DJ, Coakley KJ, et al. Comparison of breath-hold
fast spin-echo and conventional spin-echo pulse sequences for T2-weighted MR
imaging of liver lesions. Radiology1995; 194:431
-437[Abstract/Free Full Text]
- Choe KA, Smith RC, Wilkens K, et al. Motion artifact in T2-weighted
fast spin-echo images of the liver: effect on image contrast and reduction of
artifact using respiratory triggering in normal volunteers. J Magn
Reson Imaging 1997;7:298
-302[Medline]
- Low RN, Alzate GD, Shimakawa A. Motion suppression in MR imaging of
the liver: comparison of respiratory-triggered and nontriggered fast spin-echo
sequences. AJR1997; 168:225
-231[Abstract/Free Full Text]
- Kanematsu M, Hoshi H, Itoh K, et al. Focal hepatic lesion
detection: comparison of four fat-suppressed T2-weighted MR imaging pulse
sequences. Radiology1999; 211:363
-371[Abstract/Free Full Text]
- Schwartz LH, Welber A, Maier CF, et al. Fast recovery fast spin
echo evaluation of focal hepatic lesions. (abstr)
Radiology2000; 217(P):586
-587
- Augui J, Vignaux O, Argaud C, et al. Liver: T2-weighted MR imaging
with breath-hold fast-recovery compared with breath-hold half-Fourier and
non-breath-hold respiratory-triggered fast spin-echo pulse sequences.
Radiology2002; 223:853
-859[Abstract/Free Full Text]
- Katayama M, Masui T, Kobayashi S, et al. Fat-suppressed T2-weighted
MRI of the liver: comparison of respiratory-triggered fast spin-echo,
breath-hold single-shot fast spin-echo, and breath-hold fast-recovery fast
spin-echo sequences. J Magn Reson Imaging2001; 14:439
-449[Medline]
- Soyer P. Segmental anatomy of the liver: utility of a nomenclature
accepted worldwide. AJR 1993;161
: 572-573[Abstract/Free Full Text]
- Metz CE. Some practical issues of experimental design and data
analysis in radiological ROC studies. Invest Radiol1989; 24:234
-245[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
M. Bruegel, J. Gaa, S. Waldt, K. Woertler, K. Holzapfel, B. Kiefer, and E. J. Rummeny
Diagnosis of Hepatic Metastasis: Comparison of Respiration-Triggered Diffusion-Weighted Echo-Planar MRI and Five T2-Weighted Turbo Spin-Echo Sequences
Am. J. Roentgenol.,
November 1, 2008;
191(5):
1421 - 1429.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
T. Parikh, S. J. Drew, V. S. Lee, S. Wong, E. M. Hecht, J. S. Babb, and B. Taouli
Focal Liver Lesion Detection and Characterization with Diffusion-weighted MR Imaging: Comparison with Standard Breath-hold T2-weighted Imaging
Radiology,
March 1, 2008;
246(3):
812 - 822.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
