Misung Han1, Brian L Burns2, Suchandrima Banerjee2, and Janine M Lupo1,3
1Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States, 2Applications and Workflow, GE Healthcare, Menlo Park, CA, United States, 3UCSF-UC Berkeley Graduate Program in Bioengineering, University of California, San Francisco and University of California, Berkeley, San Francisco, CA, United States
Synopsis
A single scan of multi-slab, multi-echo acquisition
can simultaneously provide 3D time-of-flight (TOF) MR angiography and
susceptibility-weighted imaging (SWI) MR venography, which allows for the assessment
of vascular injury in the form of cerebral microbleeds in association with
arteries and veins. However, the acquisition for high-resolution multi-slab 3D
TOF-MRA/SWI with whole brain coverage takes over 10 minutes to acquire at 3T.
In this work, we developed a 3D multi-slab, multi-echo acquisition for TOF-MRA/SWI
with multi-band acceleration to reduce acquisition time and compared the
resulting TOF-MRA and SWI images in patients with radiation-induced microbleeds.
Introduction
Radiation therapy (RT) is a widely utilized
treatment for patients with gliomas1; however, RT can result in
collateral injury to normal-appearing brain tissue, in the formation of small chronic brain hemorrhages called cerebral
microbleeds (CMBs).2-3 The ability to assess characteristics of CMBs
in conjunction with the surrounding arterial and venous vasculature would help
to identify underlying vascular injury. A multi-slab, multi-echo acquisition to
simultaneously provide 3D time-of-flight (TOF) MR angiography (MRA) and
susceptibility-weighted imaging (SWI) MR venography has been suggested as an
effective clinical tool in quantitative evaluation of radiation-induced
vascular injury.4-6 However, the acquisition for high-resolution
multi-slab 3D TOF-MRA/SWI with whole brain coverage takes over 10 minutes to
acquire at 3T. In this work, we added multi-band (MB) acceleration7 to
a 3T multi-echo TOF-MRA/SWI sequence to reduce acquisition time and applied to
patients with radiation-induced microbleeds.Methods
Sequence: A 3D multi-slab, multi-echo sequence for TOF-MRA/SWI was
implemented for a GE 3T Signa Premier scanner (GE Healthcare, Waukesha, WI) by
modifying a commercially available 3D multi-slab, single-echo TOF sequence based
on multiple overlapping thin 3D slab acquisition
(MOTSA)8 (Figure 1). Six
additional echoes were added for SWI imaging with partial flow compensation incorporated
in the readout direction.4-5 For the simultaneous
excitation of two slabs, an original tilted optimized non-saturating ramp
excitation pulse9 was frequency-modulated to match each slab
location, and two modulated RF pulses were added together. Controlled aliasing
in parallel imaging results in higher acceleration (CAIPIRINHA) was applied by independently
modulating the phase of each RF pulse on both kz and ky
phase encoding axes, in order to provide improved g-factors for MB
reconstruction.10-11
Patients & Data Acquisition: Three patients originally diagnosed with high-grade
gliomas and had known radiation-induced microbleeds were scanned with original,
no MB-accelerated TOF-MRA/SWI sequence and MB-accelerated TOF-MRA/SWI sequence
on GE 3T Signa Premier scanner (GE
Healthcare, Waukesha, WI) using a 48-channel
phased-array brain coil. Imaging parameters used were 24x18 cm2
FOV, 384 x 288 matrix size, 1 mm slice thickness, 39 ms TR, ±41.67 kHz readout
bandwidth, TEs of 2.3/8.23/10.43/16.37/18.56/24.5/26.7 ms, 66% partial kx
readout, 20° flip angle, four axial slabs, 32 slices within each
slab, and 6 overlapping slices between adjacent slabs. In-plane acceleration by
a factor of 2 was also applied in the ky direction. Two TOF-SWI
acquisitions, without and with MB acceleration, were performed for 10 minutes and
5 minutes, respectively. An extra calibration scan was conducted before TOF-MRA/SWI
acquisitions for MB reconstruction.
Reconstruction: Image
reconstruction was conducted offline using MATLAB. For data without MB acceleration,
raw k-space data was first Fourier-transformed in the z direction, and for each
slice, in-plane ARC reconstruction12 and partial Fourier
reconstruction with a projection onto convex sets (POCS) algorithm13 were
applied. For data with MB acceleration, slice GRAPPA reconstruction was
additionally applied to unalias two superimposed slices14 after in-plane
ARC reconstruction. After reconstructing complex images for each coil, images
from the four slabs were concatenated. A composite SWI image was attained from the
2nd to 7th echo images after creating the high-pass
filtered phase image using a 2D Hanning filter and SWI image for each echo.4,6Results and Discussion
Figure 2 demonstrates
the first echo images for TOF-MRA, the maximum intensity projection (MIP) of
TOF-MRA images over the thickness of 8 mm (TOF-MRA MIP-8 mm images), composite SWI images, and the minimum
intensity projection (mIP) of composite SWI images over the thickness of 8 mm
(SWI mIP-8 mm images), from original and MB-accelerated images from one
patient. No visible aliasing artifacts were observed even though MB-accelerated
images had slightly lower SNR. From SWI mIP-8 mm images, radiation-induced CMBs
(depicted by arrows) were clearly visible on both acquisitions. To further
assess the difference in arterial and CMB contrasts from two acquisitions, line
profiles across arteries and CMBs were compared between the two acquisitions, after
resampling images with in-plane pixel size to 0.234x0.234 mm2. Figure
3 shows line profiles across six arteries over TOF-MRA MIP images (over 3.9 cm
thick inferior brain) from another patient; Figure 4 shows line profiles across
four CMBs detected from the two patients on SMI mIP-8 mm images. The two
figures illustrated that arterial and CMB contrasts from both acquisitions were
similar, indicating MB acceleration would not diminish the ability to detect
vessels and CMBs.
Figure 5 shows TOF-MRA MIP-6 cm images and SWI mIP-8mm images
from the central brain after sagittally-reformatting from both acquisitions
(patient shown in Figure 3). Arteries and veins were well-visualized with
either TOF-MRA MIP or SWI-mIP images; however, MB acceleration slightly reduced
arterial signal in the upper slabs of the TOF-MRA, probably due to insufficient
time to recover to full magnetization in flowing spins after being partially saturated while
transversing the other excited slab below. Conclusion
This work demonstrates
that MB acceleration can be combined with multi-slab 3D TOF-MRA/SWI to
reduce acquisition time without degrading depiction of vessel/CMB contrasts or yielding
visible aliasing artifacts. Future work
will further optimize the acquisition and post-processing pipelines to recover
arterial signal in the upper-most slabs and to improve overall MB-accelerated image
quality. Acknowledgements
This work was supported by a technology development
research grant from GE Healthcare.
References
1. Laperriere N, Zuraw L, Cairncross G. Radiotherapy for newly
diagnosed malignant glioma in adults: a systematic review. Radiother Oncol
2002;64:259–273.
2. Valk PE, Dillon WP. Radiation injury of the brain. AJNR Am J
Neuroradiol 1991;12(1):45-62.
3. Lupo JM, Chuang CF,
Chang SM, Barani IJ, Jimenez B, Hess CP, Nelson SJ. 7-Tesla susceptibility weighted
imaging to assess the effects of radiotherapy on normal-appearing brain in
patients with glioma. Int J Radiat Oncol Biol Phys 2012;82(3):e493–500.
4. Bian W, Banerjee S,
Kelly DA, Hess CP, Larson PE, Chang SM, Nelson SJ, Lupo JM. Simultaneous
imaging of radiation-induced cerebral microbleeds, arteries and veins, using a
multiple gradient echo sequence at 7 Tesla. J Magn Reson Imaging 2015;
42(2):269-79.
5. Du
YPP Jin ZY, Hu YZ, Tanabe J. Multi-Echo Acquisition of MR Angiography and
Venography of the Brain at 3 Tesla. J Magn Reson Imaging 2009;30(2):449-454.
6. Zou X, Bian W,
Tamir JI, Banerjee S, Chang SM, Lustig M,
Nelson SJ, and Lupo JM. Application of 3D High-resolution
Multi-echo TOF-SWI Acquisition in Radiation-induced Cerebral Microbleeds at 3T,
In:
Proceedings of the 23rd Annual Meeting of ISMRM, Toronto, Ontario, Canada, p. 66.
7. Larkman
DJ, Hajnal JV, Herlihy AH, Coutts GA, Young IR and Ehnholm G. Use of multicoil
arrays for separation of signal from multiple slices simultaneously excited. J
Magn Reson. Imaging 2001;13:313-331.
8. Parker
DL, Yuan C, Blatter DD. MR angiography by multiple thin slab 3D acquisition.
Magn Reson Med 1991;17(2):434-451.
9. Atkinson D, Brant-Zawadzki M, Gillan G, Purdy D, and Laub G. Improved
MR angiography: magnetization transfer suppression with variable flip angle
excitation and increased resolution. Radiology 1995;190(3):890-894.
10. Breuer FA, Blaimer M, Heidemann RM,
Mueller MF, Griswold MA, and Jakob PM. Controlled aliasing in parallel imaging
results in higher acceleration (CAIPIRINHA) for multi‐slice imaging. Magn Reson Med 2005;53(3):684-691.
11. Hargreaves BA, Cunningham CH, Pauly JM,
and Daniel BL. Independent phase modulation for efficient dual‐band 3D
imaging. Magn Reson Med 2007;57(4):798-780.
12. Beatty PJ, Brau AC, Chang S, Joshi SM, Michelich CR,
Bayram E, Nelson TE, Herfkens RJ, Brittain JH. A method for autocalibrating 2-D
accelerated volumetric parallel imaging with clinically practical
reconstruction times. In: Proceedings of the 15th Annual Meeting of ISMRM,
Berlin, Germany, 2007. p. 1749.
13. Haacke EM, Lindskog ED, Lin W. A Fast,
Iterative, Partial-Fourier Technique Capable of Local Phase Recovery. J Magn
Reson 1991;92(1):126-145.
14. Setsompop K, Gagoski BA,
Polimeni JR, Witzel T, Wedeen VJ, Wald LL. Blipped-controlled aliasing in
parallel imaging for simultaneous multislice echo planar imaging with reduced
g-factor penalty. Magn Reson Med 2012;67(5):1210-1224.