Ouri Cohen1,2, Frederick A. Schroeder 1,2, and Jerome L. Ackerman1,2
1Athinoula A. Martinos Center, Charlestown, MA, United States, 2Radiology, Massachusetts General Hospital, Boston, MA, United States
Synopsis
Construction of a loop-gap-resonator RF coil for MR microscopy of ex vivo mouse brains at 15T is described. The large signal obtained allowed acquisition of 20μm3 isotropic voxels, the highest resolution yet achieved in MR mouse brain imaging. Importantly, the high resolution allowed clear identification of multiple brain structures and will be of interest to researchers working with various mouse models. Introduction
Magnetic resonance microscopy is a promising
tool for high resolution 3D imaging of ex vivo soft tissue specimens since it
allows morphometric measurements and non-invasive imaging of multiple tissue
contrasts. Achieving micron-scale resolutions remains a technical challenge however due
to the rapid signal-to-noise degradation with increasing magnification. Special
consideration in RF coil design is thus required. In this work we describe the
construction of a remotely-tunable bridged loop-gap resonator RF coil optimized
for scanning perfusion-fixed mouse brains at 15T. The optimized coil allowed
acquisition of 20 μm
3 isotropic voxels that, to our knowledge,
represent the highest resolution yet achieved in MR imaging of the mouse brain.
Methods
Coaxial cylinders made of plastic (outer
cylinder) and quartz (inner cylinder), with dimensions as shown in Figure 1,
were used as formers for the coupled resonator devices. The gap in the inner
cylinder was bridged by a 24 mm long × 8 mm wide copper tape to improve the
magnetic field homogeneity [1][2]. Remote tuning was achieved by rotating the
cylinders with respect to each other and adjusting the height of the tuning
loop [3]. All experiments were conducted on a Magnex (Agilent, Yarnton, Oxford,
UK) 15 T 130 mm horizontal bore magnet equipped with a Resonance Research
(Billerica, MA, USA) 60 mm ID gradient insert and 2370 mT/m maximum
gradient interfaced to a Siemens Medical Systems (Erlangen, Germany) console. The
coil’s B
1 map was experimentally measured on a water phantom using
the manufacturer’s mapping sequence and compared to maps of a simulated coil
model designed in HFSS (Ansys, Inc., Canonsburg, PA, USA). Mouse brain
experiments were conducted, with the approval of the animal IRB committee, on freshly dissected mouse brains (Fisher Scientific,
Inc., Waltham, MA, USA) placed in sodium acetate buffered 10% formalin for a
week then transferred to a 4mM solution of Magnevist (Bayer Healthcare) in the
formalin solution for another week to shorten relaxation times. The sample was
immersed in liquid fluorocarbon (Fomblin) and degassed in a 50 Torr vacuum to
remove tissue air bubbles. The brain was scanned using a 3D GRE sequence with
the following parameters: flip angle=45°, FOV=12.4 mm
3, matrix
size=6203, TE=3.79ms, NEX=8 and TR=30ms which was the shortest TR achievable
on this system without overheating the gradient coil. The relatively long TR
resulted in a total scan time of 27 hours.
Results
Axial slices from the (a) simulated and (b) experimental
B
1 maps are shown in Figure 2. The mean of the normalized
experimental B
1 map was 0.91 ± 0.18, similar to the 0.93 ± 0.05 of
the normalized simulated map. The acquired mouse brain data is shown in Figures
3a-c for the coronal, axial and sagittal orientations. A comparable slice from
the Allen Developing Mouse Brain Atlas [4] is shown in Figure 3d for comparison
and identification of the various brain structures visible in the images. Some
of the visible structures include the corpus callosum, dentate gyrus and the
stratum lacunosum-moleculare.
Discussion
The proposed coil benefits from a homogenous excitation, as shown by the simulated
and experimental B
1 maps (Figure 2). Coil dimensions were expressly selected
to maximize the filling factor of the mouse brain specimens which yielded high
resolution images (Figure 3). Although it was first introduced nearly
thirty years ago, to our knowledge this is the first application of an LGR
based coil design at 15 T The high resolution and simple construction of
our coil allows precise measurements of distinct brain regions which is
important in studies of genetic or neurological pathologies. A limitation of
this coil design is that it is not suitable for larger samples or in vivo
imaging. The long scan time, however, is
a result of limitations in our system, not the coil, and can easily be reduced
with adequate gradient cooling.
Conclusion
We have demonstrated the utility of an RF coil that
is simple to construct and allows high resolution imaging of ex-vivo mouse
brains at ultra high fields. The voxel volumes acquired with this coil
represent the smallest yet achieved on a mouse brain.
Acknowledgements
Supported in part by the Athinoula A. Martinos Center for Biomedical Imaging and the Center for Functional Neuroimaging Technologies (funded by NIH grant P41-EB015896)References
[1] Koskinen et al, 1992; 98:576-588 [2] Ono et
al, MRM 2002; 47:415-419, [3] Mehdizadeh et al, IEEE Trans. On
Microw. Theory Tech., 1983; 31:1059-064, [4] Allen Institute for Brain Science.
Allen Developing Mouse Brain Atlas [Internet]. Available from: http://developingmouse.brain-map.org.