Half-fourier-acquisition single-shot turbo spin-echo (HASTE) MRI of the lung at 3 Tesla using parallel imaging with 32-receiver channel technology.

Original Research
Half-Fourier-Acquisition Single-Shot Turbo Spin-
Echo (HASTE) MRI of the Lung at 3 Tesla Using
Parallel Imaging With 32-Receiver Channel
Technology
Thomas Henzler, MD,1* Olaf Dietrich, PhD,2 Radko Krissak, MD,1
Tobias Wichmann, MSc,3 Titus Lanz, PhD,3 Maximilian F. Reiser, MD,2
Stefan O. Schoenberg, MD,1 and Christian Fink, MD1
Purpose: To assess the feasibility of half-Fourier-acquisi-
tion single-shot turbo spin-echo (HASTE) of the lung at 3
Tesla (T) using parallel imaging with a prototype of a 32-
channel torso array coil, and to determine the optimum
acceleration factor for the delineation of intrapulmonary
anatomy.
Materials and Methods: Nine volunteers were examined on
a 32-channel 3T MRI system using a prototype 32-channel-
torso-array-coil. HASTE-MRI of the lung was acquired at
both, end-inspiratory and end-expiratory breathhold with
parallel imaging (Generalized autocalibrating partially paral-
lel acquisitions�GRAPPA) using acceleration factors ranging
between R � 1 (TE � 42 ms) and R � 6 (TE � 16 ms). The
image quality of intrapulmonary anatomy and subjectively
perceived noise level was analyzed by two radiologists in con-
sensus. In addition quantitative measurements of the signal-
to-noise ratio (SNR) of HASTE with different acceleration fac-
tors were assessed in phantom measurements.
Results: Using an acceleration factor of R � 4 image blur-
ring was substantially reduced compared with lower accel-
eration factors resulting in sharp delineation of intrapul-
monary structures in expiratory scans. For inspiratory
scans an acceleration factor of 2 provided the best image
quality. Expiratory scans had a higher subjectively per-
ceived SNR than inspiratory scans.
Conclusion: Using optimized multi-element coil geometry
HASTE-MRI of the lung is feasible at 3T with acceleration
factors up to 4. Compared with nonaccelerated acquisi-
tions, shorter echo times and reduced image blurring are
achieved. Expiratory scanning may be favorable to com-
pensate for susceptibility associated signal loss at 3T.
Key Words: MRI; half-Fourier-acquisition single-shot
turbo spin-echo (HASTE); 3T; 32-channel phased array
coil; parallel imaging
J. Magn. Reson. Imaging 2009;30:541–546.
© 2009 Wiley-Liss, Inc.
COMPUTED TOMOGRAPHY (CT) is the first-line imaging
test for the visualization of lung disease in clinical prac-
tice. Recent publications on the potential increase of life-
time cancer risk by CT have intensified the discussion
and awareness of radiation exposure in diagnostic imag-
ing (1,2). Therefore, MRI as a radiation-free examination
may be considered as a valuable alternative for lung im-
aging in certain patient groups, for example, in patients
who will receive frequent follow-up examinations such as
immunocompromised patients with fever of unknown or-
igin or patients with cystic fibrosis (3,4).
However, MRI of the lungs remains technically chal-
lenging due to the following reasons: lung tissue has a
very low proton density, which is responsible for the low
signal-to-noise ratio (SNR) of lung parenchyma. In ad-
dition, multiple air–tissue interfaces with large mag-
netic field gradients cause susceptibility effects and in-
tervoxel phase dispersion of spins which lead to a very
low T2* relaxation time of the lung parenchyma (5). Last
but not least, the physiological motion of the heart and
the lungs cause motion artifacts which additionally re-
duces the image quality.
To overcome these limitations fast scan techniques,
which minimize motion artifacts and signal decay from
T2* and T2 effects are used. Several studies have dem-
onstrated the potential of half-Fourier-acquisition sin-
gle-shot turbo spin-echo (HASTE) for the assessment of
pulmonary nodules and infiltrates (4,6–12). Recently, it
1Department of Clinical Radiology and Nuclear Medicine, University
Medical Center Mannheim, Medical Faculty Mannheim - University of
Heidelberg, Germany.
2Josef Lissner Laboratory for Biomedical Imaging, Department of Clin-
ical Radiology, University Hospitals Grosshadern, Ludwig-Maximil-
ians-University, Munich, Germany.
3RAPID Biomedical GmbH, Rimpar, Germany.
*Address reprint requests to: T.H., Department of Clinical Radiology
and Nuclear Medicine, University Medical Center Mannheim, Medical
Faculty Mannheim - University of Heidelberg, Theodor-Kutzer-Ufer 1-3,
D-68167 Mannheim, Germany. E-mail: thomas.henzler@umm.de
Received March 13, 2009; Accepted June 10, 2009.
DOI 10.1002/jmri.21882
Published online in Wiley InterScience (www.interscience.wiley.com).
JOURNAL OF MAGNETIC RESONANCE IMAGING 30:541–546 (2009)
© 2009 Wiley-Liss, Inc. 541

was also shown that HASTE-MRI of the lungs is also
feasible at 3 Tesla (T), and that a higher lesion contrast
of nodules and infiltrates is observed at 3T when
compared with 1.5T (13). For the lungs, however, sus-
ceptibility increases at higher field strength, which may
degrade the image quality and thus reduce the effec-
tiveness of lung MRI (14).
The main features of HASTE are the radiofrequency
(RF) refocusing of the signal, resulting in a train of spin
echoes and the short image acquisition times, which
make HASTE relatively resistant to the magnetic sus-
ceptibility and motion artifacts in lung imaging. Be-
cause of the fast acquisition HASTE can be applied
under free breathing which is advantageous for pa-
tients who cannot hold the breath for long time (6).
Despite the short imaging time, the main challenge of
HASTE in lung imaging is the fast signal decay during the
echo train caused by the relatively short T2 transversal
relaxation which limits the maximum number of echoes
and the maximum length of the echo train after spin
excitation. As a single-shot technique, a problem of
HASTE is the blurring of small structures with short T2
components in phase-encoding direction. The image blur-
ring is caused by echoes that account for the same image
but are acquired with different signal intensities depend-
ing on their individual echo time (15). The introduction of
parallel imaging has brought an important approach to
speed up the image acquisition time without the need of
an improved hardware performance of the MR system.
Previous studies at 1.5T have shown that by using paral-
lel imaging with acceleration factors of 2–3 the image
resolution of HASTE can be increased in comparison to
non-accelerated acquisitions. In addition, T2-decay-re-
lated effects, such as image blurring, can be reduced (15).
The main disadvantage of parallel acquisition is the de-
crease of the SNR depending on the degree of k-space
undersampling and coil array geometry (16). Recent stud-
ies on cardiac and pulmonary MRI have shown that
higher acceleration factors of up to 6 can be achieved
without a degradation of image quality due to geometry
factor using optimized multichannel receiver coils (17,18).
Therefore, the purpose of this study was to assess the
feasibility of HASTE imaging of the lung at 3T using par-
allel imaging with a prototype of a 32-channel torso array
coil, and to determine to optimum acceleration factor for
the delineation of intrapulmonary anatomy.
MATERIALS AND METHODS
Subjects
The examinations were performed within a study pro-
tocol approved by the institutional ethics committee.
Nine healthy volunteers (five men, four women) without
any symptoms or previous medical history of chest dis-
ease were enrolled. The median age was 28.7 years (�
2.6 age range). Before undergoing MRI informed con-
sent was obtained from all volunteers.
Imaging Protocol
All exams were performed on a whole-body MR system
with a field strength of 3T (Magnetom Tim Trio, Siemens
Medical Solutions, Erlangen, Germany). The system
was equipped with 32 independent receiver channels
and a 45-mT/m gradient system. All volunteers were
examined using a prototype of a 32-channel torso array
coil (Rapid Biomedical, Rimpar, Germany). This array
consists of two electrically identical halves, an anterior
and a posterior array with 16 coil elements, each. The
coverage of each array half is 50 cm � 40 cm (xz). The
anterior array is flexible while the posterior array is
rigid and flat (19).
HASTE MRI was acquired at both, end-inspiratory and
end-expiratory breathhold using the following image pa-
rameters: bandwidth � 590 Hz/Pixel; TR � infinite (i.e.,
nonrepeated single-shot acquisition); field of view (FOV):
420 � 420 mm2; matrix: 256 � 256; slice thickness: 5
mm; voxel size: 1.6� 1.6� 5 mm3. Images were acquired
with parallel imaging, which was applied in phase-encod-
ing direction (left–right) using the generalized autocali-
brating partially parallel acquisition (GRAPPA) algorithm
(20) with 64 reference lines. The reference lines were ac-
quired before the data acquisition (“external” reference
scans) allowing thus the shortest possible echo trains and
echo times in accelerated acquisitions. The acceleration
factors ranged between R� 2 and R� 6. In addition, also
nonaccelerated images were acquired (R � 1). Depending
on the acceleration factor the TE ranged between 42 ms
(R� 1) and 16 ms (R� 6). Table 1 gives an overview of the
used sequence parameters. To ensure an identical
breathhold level for all accelerations, only three coronal
slices were acquired with each acceleration factor in a
single breathhold. The ventral slice was positioned at the
level of the trachea and the subsequent slices were sepa-
rated by a 10-mm gap.
In addition to the human exams phantom measure-
ments were performed for a quantitative SNR analysis
for all acceleration factors in a water phantom.
Data Analysis
The data analysis was performed by two board-certified
radiologists with more than 6 years experience in chest
MRI on a dedicated double monitor PACS workstation.
The readers were blinded to the used acceleration factors
and to the identity of the volunteers. In a consensus read-
ing, the different images were ranked in a side-by-side
comparison regarding the delineation of intrapulmonary
anatomical structures (pulmonary vessels, bronchial
anatomy), presence of artifacts, and subjectively per-
ceived SNR (with a ranking of “1” considered as the best
image and “6” as theworst). Themedian rank for intrapul-
Table 1
Sequence Parameters for Lung HASTE MRI at 3 Tesla
Acceleration
Factor (R)
TE
(ms)
FOV
(mm) Matrix
Slice
Thickness
(mm)
In-plane
spatial
resolution
(mm2)
1 42
2 26
3 21 420 � 420 256 � 256 5 1.6 � 1.6
4 16
5 16
6 16
542 Henzler et al.

monary anatomy, artifacts and the subjectively perceived
SNR for all acceleration factors was calculated.
The SNR of HASTE was analyzed in a water phantom
for all acceleration factors using a difference method, as
previously described (21). For this, the signal intensity
in a circular region of interest was divided by the stan-
dard deviation of the difference signal of two baseline
images in the same region.
RESULTS
The image quality for the delineation of intrapulmonary
anatomy was different between end-inspiratory and
end-expiratory scans (Figs. 1–8). For end-inspiratory
scans, an acceleration factor of R � 2 achieved the best
image quality for the delineation of intrapulmonary
anatomy with an acceptable image noise (Figs. 1, 5). In
some images the delineation of the intrapulmonary
anatomy was even better with R � 3; however, the
overall image quality with acceleration factors higher
than 2 was decreased by increasing wrapping artifacts
(Fig. 6). Images acquired with higher acceleration fac-
tors had a higher subjectively perceived SNR but were
also hampered by typical parallel imaging-related wrap-
ping artifacts (Figs. 2, 3). For end-expiratory scans the
best delineation of intrapulmonary anatomy was found
with R � 4 (Figs. 1, 7). Similar to images acquired in
inspiratory breathhold acceleration factors higher than
4 suffered from severe noise and artifacts, resulting in
nondiagnostic image quality (Fig. 2, 3, 8). In compari-
son to nonaccelerated images the image blurring was
substantially reduced with parallel imaging for both,
inspiratory and expiratory scans (Figs. 5, 7). The SNR of
the HASTE sequence, as assessed in a water phantom,
was highest without parallel imaging and steadily de-
creased with increasing acceleration factors (Fig. 4).
DISCUSSION
The results of our study indicate the feasibility of
HASTE imaging of the lung at 3T with high parallel
imaging acceleration factors by using a 32-channel coil.
In particular, we found that acceleration factors up to
R � 4 improve the delineation of intrapulmonary anat-
omy by reducing T2-related image blurring compared
with nonaccelerated MRI. The optimum acceleration
factor was dependent on the inspiratory level, that is,
higher acceleration factors were feasible at expiration,
where the increasing noise associated with higher par-
allel imaging acceleration factors was compensated by
the higher signal-intensity of deflated lung tissue.
Our findings confirm the results of previous studies
which have demonstrated the benefit of parallel imaging
Figure 1. Median rank of image quality of HASTE for the
delineation of intrapulmonary anatomy depending on inspira-
tory level and different acceleration factors.
Figure 2. Median rank of image artifacts at different inspira-
tory levels and acceleration factors.
Figure 3. Median rank of image quality of HASTE regarding
the subjectively perceived SNR level at different inspiratory
levels and acceleration factors.
Figure 4. Signal-to-noise ratio of phantom measurements
with HASTE at different acceleration factors. The SNR was
highest without parallel imaging and steadily decreased with
increasing acceleration factors.
HASTE Lung Imaging at 3T 543

for the image quality of HASTE MRI of the lungs. In a
study byHeidemann et al, the application of GRAPPAwith
acceleration factors of 2–3 resulted in a significant im-
provement in image quality over nonaccelerated acquisi-
tions (15). However, also a decreased SNR of images ac-
quired with parallel imaging was shown, which limits the
use of higher acceleration factors (R � 2) with standard
coil geometry. This is also confirmed by recent publica-
tions on HASTE MRI of the lungs, which all have used the
GRAPPA algorithm with an acceleration factor of R � 2
(3,13,22,23).
Our phantom measurements also demonstrate that the
SNR decreases substantially at higher acceleration fac-
tors. However, it is to be expected that the relative loss of
SNR is greater in these phantom measurements than in
lung MRI. The T2 relaxation time of the lung is consider-
ably shorter than of the water-filled phantom; hence, in
lung MRI, later echoes in the echo train contribute sub-
stantially less to the total SNR than in our phantom ex-
periments. Consequently, reducing the HASTE echo train
length in lung MRI by parallel imaging reduces the num-
ber of low-SNR echoes in k-space, which partially com-
pensates the SNR loss due the lower number of echoes.
Previous studies on cardiac and pulmonary MRI have
reported that acceleration factors up to R � 6 can be
applied if sophisticated multichannel coils are used. In a
study by Wintersperger et al, volunteers were examined
with SSFP CINE MRI with spatiotemporal sensitivity en-
Figure 5. Representative HASTE
MRI of a volunteer acquired at in-
spiratory breathhold with differ-
ent acceleration factors (as indi-
cated by the number in the right
upper corner). For inspiratory
scanning, the best image quality
was achieved with an accelera-
tion factor of 2. Images with
higher acceleration factors suf-
fered from increasing artifacts
and image noise, resulting in
nondiagnostic image quality.
Figure 6. Close-up of HASTE
MRI from Figure 5. The delin-
eation of intrapulmonary anat-
omy such as pulmonary ves-
sels is improved with parallel
imaging with less image blur-
ring (arrowheads); however,
with acceleration factors higher
than 2, the overall image quality
is decreased by increasingwrap-
ping artifacts (arrows) and in-
creasing noise.
544 Henzler et al.

coding (TSENSE) using a 32-element phased-array coil
array (18). Similar to our study, they observed a constant
decrease of the contrast-to-noise ratio with increasing
acceleration factors which were accompanied by reduc-
tions in subjective image quality. For CINE-MRI, a maxi-
mum acceleration factor of R � 4 was feasible for the
quantitative analysis of ejection fraction. In a study by
Nael et al, the feasibility of pulmonary MRA with GRAPPA
accelerated by a factor of 6 was evaluated (17). To com-
pensate for the lower SNR of high acceleration factors
measurements were performed at 3T with a high-relaxiv-
ity contrast agent, resulting in only mild or mild to mod-
erate image noise in all cases. Again, a significantly lower
SNR of the accelerated image data was demonstrated in
phantom experiments when compared with nonacceler-
ated MRI. For pulmonary HASTE MRI we did only observe
a benefit of parallel imaging up to a factor of R � 4 for
expiratory scanning and of R� 2 for inspiratory scanning.
For higher acceleration factors images were hampered by
significant noise and increasing parallel imaging-related
wrapping artifacts. These findings might not be general-
ized for other pulse sequences used for pulmonary MRI
(e.g., 3D GRE or TSE MRI). However, HASTE is generally
accepted as a standard imaging technique for lung imag-
ing and has been clinically evaluated for the assessment
of lung pathology such as pulmonary nodules and infil-
trates (3,6,7,9,12,13,24). Our study has several potential
limitations. Although theoretically there is a higher SNR
at 3T, higher susceptibility effects might limit this theo-
retical advantage for lung imaging. In fact, in a compara-
tive study on lung perfusion MRI, Nael et al reported a
significantly lower SNR of lung parenchyma and lower
image quality for parenchymal enhancement at 3T (17).
However, recently several studies have shown the feasi-
bility of lung MRI at 3T. In a study by Regier et al, 3D GRE
MRI and half-Fourier fast spin-echo were evaluated for
the detection of lung nodules. It was shown that the sen-
sitivity of 3D-GRE was equal to CT for lung nodules
greater than 5 mm (25). In patients with interstitial lung
disease, two studies by Lutterbey et al demonstrated the
Figure 7. Representative HASTE
MRI of a volunteer acquired at ex-
piratory breathhold with different
acceleration factors (as indicated
by the number in the right upper
corner). For expiratory scanning,
the best image quality was
achieved with an acceleration fac-
tor of 4. Images with higher accel-
eration factors suffered from in-
creasing artifacts and image
noise, resulting in nondiagnostic
image quality.
Figure 8. Close-up of HASTE
MRI from Figure 7. The delinea-
tion of intrapulmonary anat-
omy such as pulmonary ves-
sels is improved with parallel
imaging with less image blur-
ring (arrowheads); however,
with acceleration factors higher
than 4, the overall image qual-
ity is decreased by increasing
artifacts and increasing noise.
HASTE Lung Imaging at 3T 545

potential of T2-weighted TSE lung MRI at 3T for the de-
tection and determination of disease activity (26). In a
comparative study, Fink et al demonstrated that the le-
sion contrast of nodules and infiltrates increased from
1.5T to 3T (13).
In our study, end-expiratory scanning was rated fa-
vorable in combination with acceleration factors of
R�2. A potential reason for this could be that the FOV
fitted the thoracic dimension better during expiration
than during inspiration, thus causing less wrapping
artifacts. Another reason might be that the lower proton
density of end-inspiratory scans could lead to a worse
fit of the weighting factors for the GRAPPA reconstruc-
tion algorithm due to the decreased total SNR of image
and k-space data, resulting in a lower image quality.
Finally, the lower SNR of lung tissue in inspiration
increases the visibility of noise artifacts from parallel-
imaging-related noise amplification (described by the
geometry factor) (16).
A main advantages of HASTE as a single-shot sequence
technique is the opportunity to examine patients under
free breathing conditions (6), which is favorable for the
examination of small children or patients who cannot
hold their breath during the examination. As we found
that different acceleration factorswere feasible at different
breathhold levels this cannot be realized under free-
breathing conditions. A potential solution might be the
use of navigator techniques such as prospective acquisi-
tion correction (PACE), which allow the acquisition at de-
fined breathhold levels (27). Another limitation of this
study is that only healthy volunteers without lung pathol-
ogy were included. Therefore, our results might not be
valid for lung MRI of patients with lung pathology. In lung
pathology, the proton density is generally increased, re-
sulting in longer T2* times. While the T2* relaxation time
at 1.5T is approximately 7 ms (14) it increases to approx-
imately 35ms in atelectatic lung and tomore than 140ms
in lung tumor (28). Therefore, higher acceleration factors
might be feasible in lung pathology as well as in healthy
volunteers.
In conclusion, the results of our study indicate that lung
imaging with HASTE is feasible at 3T with acceleration fac-
tors up to 4 in end-expiratory scanning. End-expiratory
scanning seems to be favorable for the delineation of in-
trapulmonary anatomy and enables scanning with higher
acceleration factors (R � 4) compared with end-inspiratory
scanning (R� 2). Future studies are required to assess the
value ofHASTE imaging for thedetection of lungpathologies
using high accelerated parallel imaging.
REFERENCES
1. Brenner DJ, Hall EJ. Computed tomography—an increasing source
of radiation exposure. N Engl J Med 2007;357:2277–2284.
2. Einstein AJ, Henzlova MJ, Rajagopalan S. Estimating risk of cancer
associated with radiation exposure from 64-slice computed tomog-
raphy coronary angiography. JAMA 2007;298:317–323.
3. Puderbach M, Eichinger M, Haeselbarth J, et al. Assessment of
morphological MRI for pulmonary changes in cystic fibrosis (CF)
patients: comparison to thin-section CT and chest x-ray. Invest
Radiol 2007;42:715–725.
4. Eibel R, Herzog P, Dietrich O, et al. Pulmonary abnormalities in
immunocompromised patients: comparative detection with parallel
acquisition MR imaging and thin-section helical CT. Radiology
2006;241:880–891.
5. Marzola P, Osculati F, Sbarbati A. High field MRI in preclinical
research. Eur J Radiol 2003;48:165–170.
6. Hatabu H, Gaa J, Tadamura E, et al. MR imaging of pulmonary
parenchyma with a half-Fourier single-shot turbo spin-echo
(HASTE) sequence. Eur J Radiol 1999;29:152–159.
7. Vogt FM, Herborn CU, Hunold P, et al. HASTE MRI versus chest
radiography in the detection of pulmonary nodules: comparison
with MDCT. AJR Am J Roentgenol 2004;183:71–78.
8. Schafer JF, Vollmar J, Schick F, et al. [Detection of pulmonary
nodules with breath-hold magnetic resonance imaging in compar-
ison with computed tomography]. Rofo 2005;177:41–49.
9. Schroeder T, Ruehm SG, Debatin JF, Ladd ME, Barkhausen J,
Goehde SC. Detection of pulmonary nodules using a 2D HASTE MR
sequence: comparison with MDCT. AJR Am J Roentgenol 2005;
185:979–984.
10. Biederer J, Busse I, Grimm J, et al. [Sensitivity of MRI in detecting
alveolar infiltrates: experimental studies]. Rofo 2002;174:1033–1039.
11. Biederer J, Schoene A, Freitag S, Reuter M, Heller M. Simulated
pulmonary nodules implanted in a dedicated porcine chest phantom:
sensitivity of MR imaging for detection. Radiology 2003;227:475–483.
12. Both M, Schultze J, Reuter M, et al. Fast T1- and T2-weighted
pulmonary MR-imaging in patients with bronchial carcinoma. Eur
J Radiol 2005;53:478–488.
13. Fink C, Puderbach M, Biederer J, et al. Lung MRI at 1.5 and 3
Tesla: observer preference study and lesion contrast using five
different pulse sequences. Invest Radiol 2007;42:377–383.
14. Bergin CJ, Glover GH, Pauly JM. Lung parenchyma: magnetic
susceptibility in MR imaging. Radiology 1991;180:845–848.
15. Heidemann RM, Griswold MA, Kiefer B, et al. Resolution enhance-
ment in lung 1H imaging using parallel imaging methods. Magn
Reson Med 2003;49:391–394.
16. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P. SENSE:
sensitivity encoding for fast MRI. Magn Reson Med 1999;42:952–962.
17. Nael K, Fenchel M, Krishnam M, Finn JP, Laub G, Ruehm SG. 3.0
Tesla high spatial resolution contrast-enhanced magnetic reso-
nance angiography (CE-MRA) of the pulmonary circulation: initial
experience with a 32-channel phased array coil using a high relax-
ivity contrast agent. Invest Radiol 2007;42:392–398.
18. Wintersperger BJ, Reeder SB, Nikolaou K, et al. Cardiac CINE MR
imaging with a 32-channel cardiac coil and parallel imaging: im-
pact of acceleration factors on image quality and volumetric accu-
racy. J Magn Reson Imaging 2006;23:222–227.
19. Lanz T, Kellman P, Nittka M, Greiser A, Griswold M. A 32 channel
cardiac array optimized for parallel imaging. Proceedings of the
14th Annual Meeting of ISMRM, Seattle, 2006. (abstract 2578).
20. Griswold MA, Jakob PM, Heidemann RM, et al. Generalized auto-
calibrating partially parallel acquisitions (GRAPPA). Magn Reson
Med 2002;47:1202–1210.
21. Dietrich O, Raya JG, Reeder SB, Reiser MF, Schoenberg SO. Mea-
surement of signal-to-noise ratios in MR images: influence of mul-
tichannel coils, parallel imaging, and reconstruction filters. J Magn
Reson Imaging 2007;26:375–385.
22. Schmidt GP, Baur-Melnyk A, Herzog P, et al. High-resolution
whole-body magnetic resonance image tumor staging with the use
of parallel imaging versus dual-modality positron emission tomog-
raphy-computed tomography: experience on a 32-channel system.
Invest Radiol 2005;40:743–753.
23. Schmidt GP, Wintersperger B, Graser A, Baur-Melnyk A, Reiser
MF, Schoenberg SO. High-resolution whole-body magnetic reso-
nance imaging applications at 1.5 and 3 Tesla: a comparative
study. Invest Radiol 2007;42:449–459.
24. Bruegel M, Gaa J, Woertler K, et al. MRI of the lung: value of
different turbo spin-echo, single-shot turbo spin-echo, and 3D gra-
dient-echo pulse sequences for the detection of pulmonary metas-
tases. J Magn Reson Imaging 2007;25:73–81.
25. Regier M, Kandel S, Kaul MG, et al. Detection of small pulmonary nod-
ules inhigh-fieldMRat3T: evaluationof different pulse sequencesusing
porcine lung explants. Eur Radiol 2007;17:1341–1351.
26. LutterbeyG,GroheC,GiesekeJ, et al. Initial experiencewith lung-MRIat
3.0T: comparison with CT and clinical data in the evaluation of intersti-
tial lung disease activity. Eur J Radiol 2007;61:256–261.
27. Molinari F, Eichinger M, Risse F, et al. Navigator-triggered oxygen-
enhanced MRI with simultaneous cardiac and respiratory synchro-
nization for the assessment of interstitial lung disease. J Magn
Reson Imaging 2007;26:1523–1529.
28. Kauczor HU, Kreitner KF. MRI of the pulmonary parenchyma. Eur
Radiol 1999;9:1755–1764.
546 Henzler et al.