Christopher T Sica1, Parisa Lofti Poshtgol2, Sebastian Rupprecht3, and Qing X Yang2
1Radiology, Pennsylvania State University College of Medicine, Hershey, PA, United States, 2Neurosurgery, Pennsylvania State University College of Medicine, Hershey, PA, United States, 33HyQ Research Solutions, LLC, College Station, TX, United States
Synopsis
Prior work has explored the use of dielectric materials to enhance
imaging of the brain and spine with rectangular blocks. One area that has not
been examined to date is knee imaging, where space constraints limit the use of
blocks. In this work we evaluate the performance of an ultra-high dielectric
constant (uHDC) curved plate (εr ~ 4500) for knee imaging in a
clinical array at 3T. SNR gains of 9 to 71% percent were observed across the patella
and patellar cartilage. Transmit efficiency increased up to 100% close to the
material, with an increase in transmit inhomogeneity.
Introduction
Prior work has examined the use of uHDC materials for enhancing SNR and
reducing transmit power in several scenarios at 3T, with rectangular blocks
around the brain1 & spine2, or a conformal helmet3.
There are imaging configurations that could benefit from the application of
uHDC material but require a different geometrical package. One configuration
that has not been examined to date is that of a knee phased array. Typically
knee arrays integrate a single local transmit channel alongside a receive
phased array with 15 – 30 coil elements within a cylindrical former. Space
constraints within the coil limit the usage of rectangular uHDC blocks, where a
curved geometry would be more ideal. This work evaluates the use of a curved
uHDC plate to enhance the performance a clinical knee array at 3T.Methods
The uHDC configuration is depicted in Fig. 1. A single curved uHDC plate
was placed on top of a subject’s knee. The uHDC plate is an experimental sample
provided by HyQ Research Solutions, LLC, with dimensions of 150 x 98 x 8 mm and
a stated permittivity of 4500. The plate was nearly flush with the top half of
the knee array once the knee array is assembled (not pictured). Experiment:
A single healthy subject was scanned, with and without the uHDC plate. B1+
maps (1.83 x 1.83 mm) were collected with a 2D multi-slice pre-saturated
TurboFlash sequence4 to quantify transmit efficiency. A 2D
multi-slice small tip-angle GRE sequence (1 x 1 mm) with an integrated
noise-prescan was acquired to calculate signal-to-noise ratio with the SNR
units method5. Both sequences acquired 50 slices at 3 mm thickness. The
SNR maps were subsequently divided by the transmit sensitivity to remove the
influence of the flip angle on the measurement. Data was collected on a Siemens
3T PrismaFit (Siemens Healthineers, Erlangen, Germany) with a 15-Ch QED
transmit/receive knee array.Results
Transmit efficiency maps are displayed in Fig. 2. There is a strong
enhancement of the transmit efficiency in the region close to the plate, with
some local values reaching nearly double that of baseline. Regions that are
distant to the plate suffer from a loss of transmit efficiency relative to
baseline. Figure 3 displays a selection of SNR maps in a mosaic format. General
regions of enhancement are marked with red rectangles, and the enhancement
extends to adjacent slices. Figure 4 displays a cropped view of several slices.
Within the cropped view, significant enhancement is observed in the patellar
region up the top of the knee. Within the patella, the SNR enhancement ranged
between 45 to 71%, measured across several slices. In the patellar cartilage,
the enhancement ranged between 9.5 to 39.9%. Figure 4 displays the noise
correlation matrix. A slight increase in noise correlation was observed between
most neighboring pairs, save for the correlation between elements 6 and 7 which
has an increase of about 0.5. Dielectric material overlapping neighboring coil
elements can potentially cause strong increases in noise correlation.Discussion and Conclusion
The application of a curved uHDC plate within a clinical knee array
offers SNR increases between 9 – 71% within a region of interest close to the
material. The uHDC plate was constructed with a curvature suitable for this application
but was not designed specifically for the knee array used in this study.
Further SNR gains may be achievable with optimization of the uHDC permittivity
and geometry. Gains in SNR are potentially useful for higher-resolution imaging
of the surfaces within the knee. The B1+ mapping scans acquired
in this study used approximately the same transmit power (both with and without
the uHDC plate), which led to higher flip angles close to the dielectric (80 –
90 degrees) compared to baseline (50 degrees). When the flip angle is correctly
calibrated in a region close to the material, a power reduction of about 70%
would be expected. Additionally, there was an increase of B1+ inhomogeneity
with the uHDC plate. The gradient echo scans acquired were resilient to the B1+
inhomogeneity, though this will pose an issue for spin-echo or RARE techniques.
Careful calibration of the flip angle in a region of interest, or consideration
of the uHDC permittivity and geometry, will be necessary to overcome this
issue.Acknowledgements
We would like to acknowledge HyQ Research Solutions for providing the dielectric
material used in this study. Sebastian Rupprecht and Qing X.Yang are affiliated
with HyQ Research Solutions.References
[1] Rupprecht
et al, “Improvements of transmit efficiency and receive sensitivity with
ultrahigh dielectric constant (uHDC) ceramics at 1.5T and 3T”, MRM 2018; 79(5):
2842-51
[2] Koolstra
et al, “Improved Image Quality and Reduced Power Deposition in the Spine at 3 T
Using Extremely High Permittivity Materials”, MRM 2018; 79(5): 1192-1199
[3] Sica et al, “Toward whole‐cortex
enhancement with an ultrahigh dielectric constant helmet at 3T”, MRM 2020;
83:1123-1134
[4] Chung et al, Rapid B1+ Mapping Using a
Preconditioning RF Pulse with TurboFLASH Readout MRM“ MRM 2010; 64:439-446
[5] Kellman
et al, “Image reconstruction in SNR units: a general method for SNR measurement”,
MRM 2005; 54(6): 1439-47