close

Вход

Забыли?

вход по аккаунту

?

00004424-200512000-00002

код для вставкиСкачать
ORIGINAL ARTICLE
Magnetic Resonance Imaging of the Body Trunk Using a
Single-Slab, 3-Dimensional, T2-weighted Turbo-Spin-Echo
Sequence With High Sampling Efficiency (SPACE) for High
Spatial Resolution Imaging
Initial Clinical Experiences
Matthias Philipp Lichy, MD, MSc,* Beate M. Wietek, MD,* John P. Mugler III, PhD,储
Wilhelm Horger, MSc,¶ Marion Irene Menzel, PhD,¶ Aristotelis Anastasiadis, MD,†
Katja Siegmann, MD,* Thomas Niemeyer, MD,‡ Arnulf Königsrainer, MD,§
Berthold Kiefer, PhD,¶ Fritz Schick, MD, PhD,* Claus D. Claussen, MD,*
and Heinz-Peter Schlemmer, MD*
Purpose: The authors conducted a clinical evaluation of single-slab,
3-dimensional, T2-weighted turbo-spin-echo (TSE) with high sampling efficiency (SPACE) for high isotropic body imaging with large
field-of-view (FoV).
Materials and Methods: Fifty patients were examined in clinical
routine with SPACE (regions of interest: pelvis n ⫽ 30, lower spine
n ⫽ 12, upper spine n ⫽ 6, extremities n ⫽ 4) at 1.5 T. For achieving
a high sampling efficiency, parallel imaging, high turbofactor, and
magnetization restore pulses were used. In contrast to a conventional
TSE imaging technique with constant flip angle refocusing, the
refocusing pulse train of the SPACE sequence consists of variable
flip angle radiofrequency pulses along the echo train.
Results: Signal-to-noise ratio and contrast-to-noise ratio of SPACE
images were of sufficient diagnostic value. The possibility of image
reconstruction in multiple planes was of clinical relevance in all
cases and simplified data analysis.
Conclusion: The achievement of 3-dimensional, T2-weighted TSE
magnetic resonance imaging with isotropic and high spatial resolution and interactive 3-dimensional visualization essentially improve
the diagnostic potential of magnetic resonance imaging.
Received June 22, 2005 and accepted for publication, after revision, August
24, 2005.
From the *Departments of Diagnostic Radiology, †Urology, ‡Orthopedics,
and §General Surgery, University of Tübingen, Tübingen, Germany; the
㛳Departments of Radiology and Biomedical Engineering, University of
Virginia, Charlottesville, Virginia; and ¶Siemens Medical, Siemens AG,
MED MREA, Erlangen, Germany.
Dr. John P. Mugler was supported in part by NIH Grant R01-NS035142 from
the National Institute of Neurologic Disorders and Stroke.
Reprints: Matthias Philipp Lichy, MD, MSc, University of Tübingen, Department of Diagnostic Radiology and Section on Experimental Radiology,
Hoppe-Seyler-Stra␤e 3, 72072 Tübingen, Germany. E-mail: matthias.lichy@
med.uni-tuebingen.de.
Copyright © 2005 by Lippincott Williams & Wilkins
ISSN: 0020-9996/05/4012-0754
754
Key Words: 3D MRI, T2w, whole-body, pelvis, spine, turbospin-echo, SPACE
(Invest Radiol 2005;40: 754 –760)
M
agnetic resonance imaging (MRI) offers not only the
best soft-tissue contrast of all section imaging techniques; with the introduction of new coil architectures and
parallel imaging techniques,1–3 but this imaging technique is
now able to scan large areas of interest within one single
examination. This results in application of MRI in a clinical
context that was until today an indication for the use of
multislice computed tomography (MS CT).4 However, MS
CT is not only characterized by the acquisition of large field
of views within one examination, it can also provide real
isotropic 3-dimensional (3D) datasets with high resolution
关down to (0.4 mm3) with 64 MS CT scanners5兴. This is
especially of great relevance if multiplanar anatomic reconstructions, eg, for spine surgery preparation6,7 or tumor resection,8 are required and is essential for an improved diagnostic and therapeutic development in patient health care.
Since the introduction of MRI in clinical routines, 3D
imaging techniques were favored by radiologists and referring physicians not only because of potential reduction of
total measurement time by avoiding additional 2-dimensional
scan planes, but also this imaging technique allows retrospective free alignments of images according to anatomic/pathologic structures. A large variety of different imaging sequences are today available for this purpose3 based mainly on
gradient-echo (GRE) schemes, providing highly spatial and
time-resolved 3D datasets, eg, for contrast-enhanced MR
angiography.9 Regardless of the success of T1-weighted (w)
GRE 3D MRI, the need for a fast, highly resolved T2w 3D
MRI could not as yet be fulfilled.10 Conventional GRE imaging
Investigative Radiology • Volume 40, Number 12, December 2005
Investigative Radiology • Volume 40, Number 12, December 2005
techniques are not capable of providing a competitive T2w
contrast, whereas conventional spin-echo (SE) and turbospin-echo (TSE) schemes have not been applicable in clinical
routine as 3D schemes resulting from extremely long acquisition time required 关different 3D MR sequences based on
GRE with a T2-like contrast are available, eg, the Dual-Echo
Steady State (DESS) sequence兴. However, over the past
years, ideas were proposed to overcome this limitation of
TSE imaging.11–15 A combination of 3D T2w TSE with
variable flip angle distribution but without a restore pulse
关driven equilibrium Fourier transform technique (DEFT)兴16
was already successfully applied for imaging of the brain,11
providing outstanding isotropic resolution, competitive T2w
contrast, and acceptable acquisition time (less than 10 minutes for the whole dataset). However, for application of 3D
T2w TSE with variable flip angle distribution 关Sampling
Perfection with Application optimized Contrasts using different flip angle Evolutions (SPACE)兴 outside the brain, further
problems occur. First, tissues of the body trunk are characterized by a large variety of different relaxation times, reducing the potential of SPACE to provide a high T2w tissue
contrast and therefore to detect pathologic tissue alterations/
lesions. Second, large fields of view are required for this
purpose to visualize anatomic structures and also to avoid
severe foldin artifacts. Additionally, this is coupled with the
technical issue of reduced signal-to-noise ratio (SNR) and
inhomogeneity of coil sensitivities while acquiring large
volumes with surface phased-array coils compared with measurements with dedicated small volume coils, eg, multichannel head coils. In addition, with the combination of parallel
imaging techniques, SNR and signal homogeneity are further
reduced; these technical limitations could, therefore, hinder
the overall image and concordantly the diagnostic quality of
a highly resolved SPACE MRI of the body trunk. Finally,
motion and pulsation as well as varying patient preparation
techniques for better visualization of anatomic structures, eg,
of the rectum, have to be considered when applying such a
3D T2w TSE MR imaging technique outside the brain. The
purpose of this study was therefore to evaluate the clinical
feasibility of 3D T2w TSE (SPACE) with magnetization
restore pulse and variable flip angle distribution in wholebody applications with demand for isotropic high-resolution
T2-weighted MRI.
MATERIALS AND METHODS
Three-Dimensional Turbo-Spin-Echo
(SPACE)—Technical Considerations
The applied single-slab, 3D, TSE with restore pulse and
variable flip angle distribution (SPACE) sequence is a variant
of a TSE sequence allowing for extremely large turbo factors.
However, in contrast to a conventional TSE imaging technique with constant flip angle refocusing pulses (for instance,
180°), the refocusing pulse train of the SPACE sequence
consists of variable flip angle radiofrequency (RF) pulses
(⬍180°) along the echo train. The idea of reducing the flip
angles in TSE imaging was previously proposed in the
literature, eg, by Hennig et al17 in 1988. In their work,
constant, low flip angle refocusing RF pulses were applied for
© 2005 Lippincott Williams & Wilkins
MRI of the Body Trunk
the refocusing train. This, of course, introduces a T1 dependence of the signal and thus alters the image contrast. Nevertheless, the implementation of T1w signal contribution in
the recorded signal allows lengthening of the echo train
duration and thus the acceleration of the acquisition. To
overcome the signal evolution of the concept by Hennig et al,
Alsop et al15 proposed prescribed signal evolutions neglecting relaxation. In their concept, variable flip angles yield a
prescribed evolution of the signal during the echo train, a
“pseudo steady-state” constant signal level neglecting relaxation. Finally, Mugler et al11 considered the relaxation for
determining variable flip angle echo trains to achieve a
desired signal evolution for the tissues of interest. However,
because of the short relaxation times of the tissues of interest
of the body trunk, SPACE in our examinations is used in a
pseudo-steady-state mode. Signal evolution can be considered largely as identical with conventional TSE scheme, eg,
using constant 150° flip angles for reduced SAR.
Magnetic Resonance Unit and Sequence
Parameter
All examinations were performed using a single 1.5-T
whole-body MR scanner (Magnetom Avanto; Siemens Medical Solutions, Germany) equipped with 32 independent RF
channels. In all cases, a combination of body and spine matrix
coils (in addition, use of combination of neck and head
phased-array coils for imaging the cervical spine) was used
for signal reception 关total imaging matrix (TIM) system兴. The
MR unit features gradient fields strength of maximum 45
mT/m (72 mT/m effective) and a slew rate of maximum 200
T/m/s (346 T/m/s effective). The potential field of view is set
to 50 cm maximum.
Sequence parameters for 3D SPACE (slice selective,
work in progress sequence; Siemens Medical) were: TR/TE ⫽
1500/124 ms, 2 averages, field of view (FoV) ⫽ 380 mm, 144
slices per slab, base matrix ⫽ 384, iPAT factor ⫽ 3 (GRAPPA,
24 reference lines, triple matrix coil mode), bandwidth 407
Hz/Px, turbo factor ⫽ 71, 2 echo trains per partition, echospacing ⫽ 4.2 ms, resulting isotropic voxel size (1.0 mm3)
(no interpolation; prescan normalize raw filter; 6/8 slice
partial Fourier), resulting in a measurement time of 10 minutes 32 seconds (80% FoV phase) or 13 minutes 26 seconds
(100% FoV phase). Slab orientation was coronal or sagittal.
The number of slices could be adapted to the required slab
thickness (range up to 244 slices). In case of a small volume
of interest, minimum measurement time was approximately 7
minutes in our study. No saturation pulses and/or fat suppression techniques were applied.
For conventional 2D T2w TSE, in n ⫽ 3/2/3 MR
examinations (coronal/transversal/sagittal orientation), parallel imaging techniques were used, also (iPAT factor of 2;
Grappa; dual coil mode). No respiratory triggering techniques
were used.
Patient Handling
All patients were referred consecutively to our MR
unit. There were no special inclusion or exclusion criteria
despite contraindications for MR examinations. All examinations were performed while the patients were breathing
755
Lichy et al
Investigative Radiology • Volume 40, Number 12, December 2005
freely. For imaging of rectal cancer, a dilution of water and
ultrasound gel additionally was used. If required, disturbing
bowel motion was reduced by applying n-butyl-scopolamine
(Buscopan) intravenously. The attending physician was allowed to resign conventional nonfat-suppressed 2D T2w TSE
sequences in case of sufficient diagnostic image quality of
SPACE and if conventional nonfat-suppressed 2D TSE was
mandatory according to clinical MRI sequence protocol. All
patients were informed in detail about the matter and purpose
of the MR examination before giving their written consent.
Conduction and reporting of the MR examinations were in
accordance with standard clinical routine procedures and
were in concordance with the guidelines of the Declaration of
Helsinki. Application of non-U.S. Food and Drug Administration-approved MR sequences for patient care is covered by
national regulations.
Patients
A total of 50 patients were examined (21 female and 29
male) with SPACE (nonfat-suppressed). Mean age was 49
years (range, 2–79 years). Anatomic regions of interests
were: pelvis n ⫽ 30, lower spine n ⫽ 12, upper spine n ⫽ 6,
and extremities n ⫽ 4. In 10 cases, rectal filling with ultrasound gel was performed (staging of rectal cancer). For
whole spine imaging, the automated table-moving system of
the MR scanner was used without repositioning of the patient
and applied coils (2 patients; application of 2 SPACE slabs).
Composition of these images as well as multiplanar reconstruction of SPACE images and interactive reporting of 3D
datasets were all performed using the commercially available
scanner software (Syngo MR 2004V; Siemens Medical).
Magnetic Resonance Data Evaluation
Evaluation of SPACE MRI data included axial, sagittal,
and coronal reconstructions 关multiplanar reconstruction
(MPR); 1-mm slice thickness兴. Appraisal of the resulting
images was performed by 2 radiologists (first reader 3 years
MR experience, second reader 13 years MR experience;
consensus read). Documentation of the rating was performed
by 5-point ordinal scale (1 ⫽ poor/nondiagnostic, 2 ⫽ below
average, 3 ⫽ fair, 4 ⫽ good, 5 ⫽ excellent). Technical
evaluation of SPACE images included: overall SNR impression, including overall signal homogeneity (fatty tissue and
background signal), overall contrast-ratio impression, and
presence of artifacts (if applicable, including description of
artifact and its influence on diagnostic potential). Criteria for
subjective evaluation of contrast-to-noise ratio (CNR) were
signal provided by liquid structures (bladder, liquor), signal
differences of muscles, fat, and bone marrow and, if applicable, contrast of pathologic structures/surrounding tissues.
In the case of pathologic structures/signal behavior, 3D
SPACE was compared with conventional 2D T2w images
(with or without fat suppression) for completeness of visualization of these areas. Additionally, diagnostic quality of
SPACE and, if applicable, conventional T2w nonfat-suppressed MRI was rated (statistical analysis by paired 2-sided
Student t test with P ⬍ 0.05 as significance level; data
analysis was performed with JMP version 1.5.2, SAS Institute Inc.). Use of a 3D evaluation tool for diagnosis was also
756
rated with the SPACE sequence (2 points: 0 ⫽ no additional
information, 1 ⫽ additional information/higher diagnostic
confidence).
For evaluation of the MR examinations within a clinical
setting, the following points were reported: total examination
time, number of sequences, application of contrast mediaenhanced (ce) MRI, extent of anatomic region covered by
conventional T2w MRI, as well as sequence parameters of
T2w sequences.
RESULTS
In 32 cases, conventional nonsuppressed T2w imaging
was available for direct comparison of diagnostic potential of
SPACE and conventional 2D T2w MR imaging. In 26 patients, MRI included 2D T2w fat-suppressed MRI (TSE) and
in 16 patients fat as well as nonfat-suppressed T2w imaging
was available. An overview of nonfat-suppressed 2D T2w
MR sequence parameters is given in Table 1. The highest
inplane resolution of all conventional 2D T2w MRI was 0.7 ⫻
0.5 mm2. In this patient, a partial rupture of the ligamenta
alaria was suspected and therefore an MRI of the upper
cervical spine was conducted. However, edema within these
ligaments was clearly visualized only by SPACE in the
combination with interactive 3D MPR software.
N-butyl-scopolamine was applied routinely in all cases
of imaging the pelvis before T2w MRI. No significant differences between SPACE and conventional 2D T2w TSE
MRI was observed concerning artifacts caused by bowel
movement.
All lesions/pathologies detected by 2D T2W MRI (nonas well as fat-suppressed MRI) were visualized by SPACE. In
one case, with coincidental finding of large kidney cysts,
SPACE was able to visualize the extent of these lesions, but
signal behavior was only slightly hyperintense compared with
the surrounding fatty tissue. In this case of a pelvic MRI
examination, these findings were detected as a result of the
large applied FoV of the SPACE sequence and only comparable with coronal T2w HASTE MRI.
Detailed information about SNR, CNR, and diagnostic
rating is given in Table 2. SNR and CNR ratings for SPACE
were high. In 2 cases, SNR of SPACE was rated as poor but
still diagnostic. In both cases, technical problems with one
coil element of the body matrix were present, affecting the
whole MR examination, including conventional 2D T2w
MRI. CNR of coronal orientation was rated significantly
higher (coronal vs transversal MPR P ⫽ 0.04; coronal vs
sagittal MPR P ⫽ 0.01) compared with transversal or sagittal
reformation (primary slab orientation of SPACE was in 48
cases coronal, in 4 cases sagittal). In all these cases, images
appeared to be “blurring,” but without affect on diagnostic
image quality. In cases of possible direct comparison between
SPACE and conventional 2D T2w, diagnostic quality of
SPACE was considered as equivalent or superior to conventional 2D T2w MRI. Ratings of diagnostic quality of SPACE
were slightly higher compared with conventional nonfatsuppressed T2w TSE in coronal orientation (SPACE vs
2D T2w: coronal P ⬍ 0.01; transverse P ⫽ 0.05; sagittal
P ⫽ 0.17).
© 2005 Lippincott Williams & Wilkins
Investigative Radiology • Volume 40, Number 12, December 2005
MRI of the Body Trunk
TABLE 1. Overview of the Applied Magnetic Resonance Sequence Parameters of All SPACE and Conventional 2-Dimensional
(2D) T2-Weighted Turbo-Spin-Echo (TSE) Examinations
Sequence Parameters
(mean ⴞ SD)
(minimum/maximum/
median)
2D T2w TSE
SPACE
Coronal/Sagittal
TR (ms)
1500
TE (ms)
124
Slab thickness (mm;
no. of slices plus
distance factor)
Field of view (mm)
146 ⫾ 23
88/192/142
Acquisition time (s)
650 ⫾ 99
432/1040/631
380 * 380
1
Resulting voxel size
(mm3)
Axial
Coronal
Sagittal
6040 ⫾ 1694
3800/8750/5985
113 ⫾ 14
92/134/122
207 ⫾ 96
49/488/225
5942 ⫾ 1240
3801/8230/5905
115 ⫾ 10
96/122/122
164 ⫾ 45
99/238/174
6199 ⫾ 1291
4370/9170/6050
117 ⫾ 9
99/134/122
100 ⫾ 6
99/129/99
(227 ⫾ 36) * (257 ⫾ 52)
(154/280/228) * (160/350/260)
261 ⫾ 95
(294 ⫾ 54) * (294 ⫾ 54)
(220/430/280) * (220/430/280)
233 ⫾ 63
(299 ⫾ 69) ⫾ (307 ⫾ 62)
(195/400/280) * (220/400/280)
196 ⫾ 11
115/546/250
2.35 ⫾ 0.85
80/330/241
3.59 ⫾ 4.25
145/199/199
2.37 ⫾ 0.09
1/5.10/2.06
1.64/17.62/2.39
2.06/2.39/2.39
All examinations were performed according to clinical routine magnetic resonance imaging protocol and sequence parameters were adapted according to clinical needs/question.
Use of interactive a 3D MPR software tool for interpreting SPACE 3D datasets was considered to be helpful and
improved diagnostic confidence in all cases. Only SPACE
datasets and interactive 3D MPR allowed planning of instruments in cases of spinal surgery. For imaging of rectal cancer,
SPACE was also able to properly visualize the borders of the
mesorectum. Furthermore, better differentiation between
small vessels and lymph nodes was possible when applying
SPACE and interactive MPR (Fig. 1).
TABLE 2. Rating of 2 Experienced Radiologists (Consensus)
Concerning Signal-to-Noise Ratio (SNR), Contrast-to-Noise
Ratio (CNR), and Diagnostic Quality of SPACE and
Conventional 2-Dimensional (2D) T2-Weighted (T2w) TurboSpin-Echo Magnetic Resonance Imaging (5-Point Scale With
1 ⫽ Poor/Nondiagnostic and 5 ⫽ Excellent)
SNR
CNR
Diagnostic Quality
4.2 ⫾ 0.7
2/5/4
4.0 ⫾ 0.7
2/5/4
4.1 ⫾ 0.6
2/5/4
4.1 ⫾ 0.6
3/5/4
3.9 ⫾ 0.7
3/5/4
3.8 ⫾ 0.6
3/5/4
4.2 ⫾ 0.7
2/5/4
4.1 ⫾ 0.7
2/5/4
4.0 ⫾ 0.8
2/5/4
NA
NA
Transverse
NA
NA
Sagittal
NA
NA
3.2 ⫾ 0.8
2/4/3
3.8 ⫾ 0.8
2/5/4
3.7 ⫾ 0.7
2/5/4
SPACE
Coronal
Transverse
Sagittal
2D T2w
Coronal
NA, Not applicable.
© 2005 Lippincott Williams & Wilkins
DISCUSSION
Assessment of highly resolved anatomic data within a
short examination time is essential for in-depth understanding
of the pathologic process.3,6,8 Especially in the case of surgical treatment planning, conventional 2D MRI is often not
capable of visualizing all anatomic structures required for
optimal preparation of surgical intervention (eg, the mesorectal fascia and anatomic variations of the rectal muscles in case
of resection of rectal cancer). In addition, such detailed
anatomic information is worthless if information regarding
extension of a disease (eg, tumor infiltration) is missing.
Recently, MS CT scanners have been superior in providing
isotropic 3D datasets; however, this method is often hampered by its inherent and insufficient contrast behavior for
evaluating soft tissue lesions. It is, therefore, of great importance to combine the information given by all these different
modalities and convert the given specific data into a treatment
concept. In clinical routine, this results in multiple MRI and
CT scans to benefit of both techniques, but this procedure
often results in treatment delays and additional costs. To
advance image interpretation and to reduce loss of information, image fusion techniques are used.18,19 As an example,
the fusion of positron emission tomography (PET) and CT
scans is now considered to be superior to the single PET and
CT examinations.20 –22 However, such imaging fusion techniques are time-consuming and technically challenging.23,24
The used SPACE sequence combines highly resolved
(voxel size in our study only 1 mm3) 3D datasets with high
soft tissue contrast (T2w) and is well within a tolerable
examination time for clinical routine (less than 20 minutes for
whole spine imaging; compare Fig. 2). As a variant of a TSE
sequence, based on our results, contrast behavior of SPACE
has to be regarded as sufficient for detection of T2w hyperintensities, eg, edema or tumor infiltration of muscles, regard-
757
Lichy et al
Investigative Radiology • Volume 40, Number 12, December 2005
less of examined anatomic region. With the introduction of
image acquisition with low flip, slight T1-time dependence of
tissue contrast cannot be excluded but should not seriously
hamper image interpretation. A variation of the applied
SPACE sequence using the potential of variable flip angle
evolution but without a magnetization restore pulse was
already clinically introduced in brain imaging.11 For this purpose, however, only relaxation times and signal evolution of
mainly 3 compartments (gray and white matter, cerebral
fluid) have to be taken into account; this is clearly an
advantage over abdominal and musculoskeletal imaging in
which much more tissue types have to be taken into consideration. Sequence parameters as well as flip angle evolution
were chosen for musculoskeletal and pelvic imaging in our
examinations according to the literature.12–16,25 However,
contrast and signal behavior of the liver and the kidney are
clearly different compared with conventional 2D T2w TSE
MRI and should, therefore, not be considered at present for
diagnosis within our sequence setting. As theoretical considerations concerning the signal evolution of SPACE and conventional T2w TSE MR sequences suspected, the observed
contrast of SPACE is primarily slightly different from T2w
TSE as a result of the short TR in combination with the
restore pulse, not as a result of variable flip angles. However,
the contrast of SPACE is directly comparable to conventional
2D T2w TSE MRI with restore pulse. Furthermore, the
SPACE imaging technique is sensitive to B1 homogeneity
and the radiofrequency distribution, which determines the
calculated flip angles, with direct impact on contrast and
signal characteristics. This is, however, not a specific prob-
FIGURE 1. Examples for magnetic resonance examinations
with the applied 3-dimensional (3D) turbo-spin-echo sequence (SPACE). In all images shown, spatial resolution is 1
mm3. A, Screenshot of the 3D evaluation tool used for diagnosis. A patient with a chordoma is displayed. Original data
was acquired in coronal orientation (measurement time
758
approximately 10 minutes). The thin lines in each image
represent the orientation and angulations as well as viewing
direction (small arrows) of the referring multiplanar reconstructions (MPRs). B–D, Coronal (D, original slab orientation), reconstructed sagittal (E), and axial (F) image in a case
of preoperative staging of a rectal carcinoma. For this purpose, a dilution of ultrasound gel and water was administered. To reduce bowel motion, N-butyl-scopolamine was
applied intravenously. The exophytic configured tumor is
well defined and extent/separation toward surrounding tissue is of high diagnostic quality. Evaluation of lymph nodes
up to the aortic branch is also available as a result of the
large field of view (380 mm2). E–G, A 35-year-old male patient with an extensive rhabdomyosarcoma, who was referred to our department for staging. The tumor extends to
the bladder (B) and large parts of the right pelvis, including
osseous infiltration. Additionally, diffuse tumor spread within
the muscles is obvious. Highly resolved 3D SPACE (resolution
1 mm3, 176 slices, measurement time: 12 minutes 31 seconds) was acquired before application of contrast media (E,
original slab orientation). Axial T2-weighted 2-dimensional
MPR based on SPACE data is given in F (1-mm slice thickness). However, 3D SPACE enables free orientation within
the measured volume, allowing detailed assessment of areas
of interest. For example, a diffuse infiltration of the M. obturatorius is visible (marked with *) on image G; however, reconstructed MPRs given in G and H are clearly superior for
detailed evaluation.
© 2005 Lippincott Williams & Wilkins
Investigative Radiology • Volume 40, Number 12, December 2005
FIGURE 2. A 14-year-old female patient with severe scoliosis,
Arnold-Chiari malformation, and multiple segmentation disorders of the spine. Original SPACE images are given in A.
For detailed assessment of the multiple malformation of the
vertebral bodies, an oblique reformation along the spine axis
was performed (B) in which A represents the cervical and B
the thoracic spine segment. The isotropic 3-dimensional
SPACE dataset could also provide detailed information about
the present cervical cele and the multiple malformations of
the brain (C and D).
lem of SPACE because it is also true for conventional MRI
sequences.26 –28
Despite these theoretical considerations, the applied
SPACE sequence was able to visualize all lesions detected by
conventional 2D T2w MRI. It may theoretically be possible
that thicker slices and higher in-plane resolution could provide a better visualization of anatomic details; in general, partial
volume effects of anatomic structures are of course reduced in
smaller voxel sizes and therefore the exact separation of anatomic structures is directly linked toward the total voxel volume.
In addition, SNR and CNR impression ratings for
SPACE were comparable with conventional T2w MRI. It
should be taken into account that these ratings were performed with the high-resolved MR SPACE images. In only
one case, voxel size of conventional 2D MRI was identical
with SPACE; in all the other cases, SNR of 2D T2w MRI was
raised automatically by its larger voxel sizes. It is all the more
demonstrated that a combination of high turbo factors (in our
application factor of 71), parallel imaging techniques (we
applied Grappa with iPAT factor of 3) and higher resolution
© 2005 Lippincott Williams & Wilkins
MRI of the Body Trunk
compared with conventional T2w MRI sequences can be
highly competitive in its diagnostic quality (or even superior).
Based on our data, it is also highly assumable that the relative
long measurement time required for acquisition of these 3D
datasets is tolerable, because large areas of interest have to be
covered by MRI. This is also true if multiplanar 2D T2w MRI
is required for an accurate diagnosis. In all of our cases, no
conventional 2D T2w MRI without fat suppression was
essential for diagnosis, but SPACE was in some cases.
However, it should be mentioned that the demonstrated
results do not allow conclusion that SPACE is generally
superior to 2D T2w TSE MRI. The slightly different contrast
compared with conventional could, eg, hamper the detection
of prostate cancer. Therefore, the clinical value of SPACE for
dedicated clinical questions has to be investigated in future
studies.
In conclusion, this imaging technique allows combining real T2w TSE MRI with 3D imaging within clinical
routine. For imaging of complex anatomic relationships with
the additional demand of large anatomic regions, especially
in cases of planning of spine surgery or resection of musculoskeletal tumors, the SPACE imaging technique is now
exclusively demanded by our referring physicians.
In combination with ultra-high-field MR scanner (3 T
and above), further improvements of resolution and/or measurement times can be expected. However, future efforts
should also focus on optimizing sequence parameters for
dedicated imaging of the upper abdomen, including the liver
and pancreas. For this purpose, however, breath-gating techniques (eg, PACE) have to be implemented.
REFERENCES
1. Lauenstein TC, Goehde SC, Herborn CU, et al. Whole-body MR
imaging: evaluation of patients for metastases. Radiology. 2004;233:
139 –148.
2. Schlemmer HP, Schafer J, Pfannenberg C, et al. Fast whole-body
assessment of metastatic disease using a novel magnetic resonance
imaging system: initial experiences. Invest Radiol. 2005;40:64 –71.
3. Keogan MT, Edelman RR. Technologic advances in abdominal MR
imaging. Radiology. 2001;220:310 –320.
4. Walker R, Kessar P, Blanchard R, et al. Turbo STIR magnetic resonance
imaging as a whole-body screening tool for metastases in patients with
breast carcinoma: preliminary clinical experience. J Magn Reson Imaging. 2000;11:343–350.
5. Flohr T, Stierstorfer K, Raupach R, et al. Performance evaluation of a
64-slice CT system with z-flying focal spot. Rofo. 2004;176:1803–1810.
6. Inoue M, Minami S, Nakata Y, et al. Preoperative MRI analysis of
patients with idiopathic scoliosis: a prospective study. Spine. 2005;30:
108 –114.
7. Bullmann V, Halm HF, Niemeyer T, et al. Dual-rod correction and
instrumentation of idiopathic scoliosis with the Halm-Zielke instrumentation. Spine. 2003;28:1306 –1313.
8. Akasu T, Iinuma G, Fujita T, et al. Thin-section MRI with a phasedarray coil for preoperative evaluation of pelvic anatomy and tumor
extent in patients with rectal cancer. AJR Am J Roentgenol. 2005;184:
531–538.
9. Spuentrup E, Katoh M, Buecker A, et al. Free-breathing 3D steady-state
free precession coronary MR angiography with radial k-space sampling:
comparison with cartesian k-space sampling and cartesian gradient-echo
coronary MR angiography—pilot study. Radiology. 2004;231:581–586.
10. Fautz HP, Leupold J, Hennig J, et al. Signal behavior in continuously
ramped 2D TrueFISP for whole-body imaging. Magn Reson Med.
2002;48:1085–1090.
11. Mugler JP 3rd, Bao S, Mulkern RV, et al. Optimized single-slab
759
Lichy et al
12.
13.
14.
15.
16.
17.
18.
19.
Investigative Radiology • Volume 40, Number 12, December 2005
three-dimensional spin-echo MR imaging of the brain. Radiology. 2000;
216:891– 899.
Mugler JP, Wald LL, Brookeman JR. T2-Weighted 3D Spin-Echo Train
Imaging of the Brain at 3 Tesla: Reduced Power Deposition Using Low
Flip-Angle Refocusing RF Pulses. Proceedings of the International
Society for Magnetic Resonance in Medicine 9th Meeting; Glasgow;
2001:438.
Mugler JP III, Meyer H, Kiefer B. Practical Implementation of Optimized Tissue-Specific Prescribed Signal Evolutions for Improved TurboSpin-Echo Imaging. Proceedings of the International Society for Magnetic Resonance in Medicine 11th Meeting; Toronto; 2003:203.
Mugler JP, Kiefer B, Brookeman JR. Three-Dimensional T2-Weighted
Imaging of the Brain Using Very Long Spin-Echo Trains. Proceedings of
the International Society for Magnetic Resonance in Medicine 8th
Meeting; Denver; 2000:687.
Alsop DC. The sensitivity of low flip angle RARE imaging. Magn Reson
Med. 1997;37:176 –184.
van Uijen CM, den Boef JH. Driven-equilibrium radiofrequency pulses
in NMR imaging. Magn Reson Med. 1984;1:502–507.
Hennig J. Multiecho imaging sequences with low refocusing flip angles.
J Magn Reson. 1988;78:397– 407.
Polo A, Cattani F, Vavassori A, et al. MR and CT image fusion for
postimplant analysis in permanent prostate seed implants. Int J Radiat
Oncol Biol Phys. 2004;60:1572–1579.
Lemke AJ, Niehues SM, Amthauer H, et al. Clinical use of digital
retrospective image fusion of CT, MRI, FDG-PET and SPECT—fields
of indications and results. Rofo. 2004;176:1811–1818.
760
20. Kostakoglu L, Hardoff R, Mirtcheva R, et al. PET-CT fusion imaging in
differentiating physiologic from pathologic FDG uptake. Radiographics.
2004;24:1411–1431.
21. Pfannenberg AC, Eschmann SM, Horger M, et al. Benefit of anatomicalfunctional image fusion in the diagnostic work-up of neuroendocrine
neoplasms. Eur J Nucl Med Mol Imaging. 2003;30:835– 843.
22. Bar-Shalom R, Yefremov N, Guralnik L, et al. Clinical performance of
PET/CT in evaluation of cancer: additional value for diagnostic imaging
and patient management. J Nucl Med. 2003;44:1200 –1209.
23. Romer W, Chung M, Chan A, et al. Single-detector helical CT in
PET-CT: assessment of image quality. AJR Am J Roentgenol. 2004;182:
1571–1577.
24. Townsend DW, Beyer T, Blodgett TM. PET/CT scanners: a hardware
approach to image fusion. Semin Nucl Med. 2003;33:193–204.
25. de Bazelaire CM, Duhamel GD, Rofsky NM, et al. MR imaging
relaxation times of abdominal and pelvic tissues measured in vivo at 3.0
T: preliminary results. Radiology. 2004;230:652– 659.
26. Truong TK, Clymer BD, Chakeres DW, et al. Three-dimensional numerical simulations of susceptibility-induced magnetic field inhomogeneities in the human head. Magn Reson Imaging. 2002;20:759 –770.
27. He G, Evalappan SP, Hirata H, et al. Mapping of the B1 field distribution
of a surface coil resonator using EPR imaging. Magn Reson Med.
2002;48:1057–1062.
28. Ibrahim TS, Lee R, Abduljalil AM, et al. Dielectric resonances and B(1)
field inhomogeneity in UHFMRI: computational analysis and experimental findings. Magn Reson Imaging. 2001;19:219 –226.
© 2005 Lippincott Williams & Wilkins
Документ
Категория
Без категории
Просмотров
0
Размер файла
670 Кб
Теги
00002, 00004424, 200512000
1/--страниц
Пожаловаться на содержимое документа