Natalia Gudino1, Qi Duan1,2, Jacco A de Zwart1, Stephen J Dodd1, Joe Murphy-Boesch1, Peter van Gelderen1, and Jeff H Duyn1
1Advanced MRI Section, LFMI, NINDS, National Institutes of Health, Bethesda, MD, United States, 2Food and Drug Administration, Silver Spring, MD, United States
Synopsis
On-coil current-source switch-mode amplification presents
high power efficiency, allows direct control of the transmit field (B1),
and decoupling of elements through the amplifier output impedance. These are important advantages over conventional
remote voltage-mode quasilinear amplification that should
allow more efficient and safer implementation of a multi-channel transmit
system. Following this approach, we present an
optically controlled transceiver design that was used for initial safety
assessment of the technology toward the implementation of a high channel-count pTx
array for brain imaging at high-field. We acquired human brain images with this technology
at 7T
Introduction
On-coil current-source switch-mode
amplification presents several advantages over conventional remote voltage-mode
quasilinear amplification, namely high power efficiency, optical and direct
control of the transmit field (B1), and decoupling of elements through the
amplifier output impedance1,2.
These advantages should allow more efficient and safer implementation of a
multi-channel transmit system, and a significant simplification of the coil array.
Based on our previous -300 MHz- amplifier design and pTx interface2, we built a transceiver
array by adding a miniaturized transmit-receive (TR) switch and low-noise
amplifier (LNA) to the on-coil configuration. A transceiver setup simplifies a
first safety assessment toward imaging humans by eliminating the need of
additional receive hardware. This early, if not first application of on-coil
transceivers for MRI of in-vivo human brain was performed at low transmit
power.Methods
Hardware : Optically controlled on-coil amplifiers for 7T imaging2 were connected to 6 cm
diameter loops through a 4.2 x 4.6 cm 5-layer double sided PCB that contains a
miniaturized TR switch and an in-house LNA (Figure
1). The switch provides a balanced connection from the amplifier to the coil trough a pair of low loss PIN diodes and a balanced to
unbalanced connection to the LNA through a 50 Ω lattice balun. Additional
isolation was achieved by a PIN diode in series with the LNA. During transmission
a pin diode shorts the balun output and high impedance is seen from the coil
into the RX due to the double parallel resonance formed by the Cc-Lb pair.
During signal reception the impedance transformation of the balun results in
optimum impedance seen from the LNA into the coil (around 50 Ω) and element
decoupling is provided by the low preamplifier input impedance (< 2 Ω) .
In this
preliminary work 2 channels were assembled on the outside of a half cylindrical
former to electrically insulate all electronics from the volunteer while also providing
head support (Figure 2).
Phantom Imaging: Axial images of an
oil phantom were acquired with a single transceiver element using a
gradient-echo sequence with a rectangular RF pulse, TRF=4 ms, TR=2 s, TE= 3.18
ms, FOV=144 mm x 108 mm, resolution= 1.1 mm x 1.1 mm x 2 mm. SNR maps were calculated from these images. SAR simulations: Electromagnetic simulations
were performed for two loops located at an azimuthal angle of 45 degrees and
3.7 cm separation, as assembled on the cylindrical former. Both
elements were located 2.6 cm from a cylindrical phantom (f=16 cm and l=20 cm)
that simulates brain tissue (σ=0.55 S/m and ε=52)3. The total average
input power for computing the 10-g average SAR was limited to 0.4 W. SAR was
computed while sweeping the phase of one channel in 45-degree steps. In-vivo brain imaging: The setup is shown in Figure 3.
The optical interface and amplifier design have been detailed in our previous work2. The RX signal is sent
to the scanner console through the coil connection on the patient
table. This also provides the PIN diode signals and DC bias for the LNAs. In
this initial experiment, peak power per amplifier was below 20 W by limiting
the total DC input power to 40 W. Excitation duty cycle was limited to 1% to
keep average total power below 0.4 W as it was limited for SAR simulations.
Additional safety protections were added by monitoring DC input current and fuses
in the power line. Images of a volunteer brain were acquired with a
gradient-echo sequence, with a flip angl=10 degrees, TE/TR=7/500 ms,
slice thickness =5 mm, FOV= 195 X 240 mm and matrix size= 286 X 352. Imaging
was approved by the Institutional Review Board.Results
The TR switch provided 50 dB
isolation to protect the LNA during transmission. Maximum SNR was above 400 for 80-degree nominal flip angle as shown
in Figure 4. Noise images were acquired (RF amplitude =0) with and without the
amplifier inside the bore. Less than 1% increase in the noise standard
deviation was detected. The maximum local SAR10g obtained from simulations was at
least 7 times lower than the FDA limit of 10 W/kg. Figure 5 shows a volunteer’s
occipital brain region obtained by two-channel transmission and reception.Discussion
Successful practical high field
implementation of on-coil transceive technology was demonstrated on human
brain. Our on-coil transceiver setup allowed to simplify the safety assessment
and is easily extendable to a larger number of channels. We are currently
working on a four-channel array that includes real-time monitoring of the
transmit signal4 to
achieve higher performance and ensure safety under high power operation.Acknowledgements
No acknowledgement found.References
1- Gudino N, Heilman JA, Riffe MJ, Heid
O, Vester M, Griswold MA.Magn Reson Med. 2013
Jul;70(1):276-89.
2-Gudino N, Duan Q, de
Zwart JA, Murphy-Boesch J, Dodd SJ, Merkle H, van Gelderen P, Duyn JH.Magn Reson Med. 2016 Jul;76(1):340-9.
3-Duan Q, Duyn JH, Gudino N, de Zwart JA, van Gelderen P,
Sodickson DK, Brown R. Med Phys. 2014
Oct;41(10):102303. doi: 10.1118/1.4895823.
4- Gudino N,
de Zwart Jacco A., Duan Qi, van Gelderen Peter,
and Duyn Jeff H. Proc. Intl. Soc. Mag. Reson. Med 24, 2016 (Abstract 2181)