Ehsan Kazemivalipour1,2, Bastien Guerin1,2, and Lawrence L. Wald1,2,3
1A. A. Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital, Charlestown, MA, United States, 2Harvard Medical school, Boston, MA, United States, 3Harvard-MIT Division of Health Sciences Technology, Cambridge, MA, United States
Synopsis
Keywords: Safety, Safety
Far-field electromagnetic
radiation patterns and levels were simulated on the 10m radius regulatory sphere
for conventional MRI scanners at 0.5T, 1.5T, 3T, and 7T operated without an RF
shielded room. The levels and patterns were strongly affected by the symmetry
of the load. With a body load, the peak E-fields on a 10m radius surface rose
with roughly the square of the frequency and far exceeded regulatory limits
even for 0.5T. With the body in the bore, the radiated patterns also take on a
highly asymmetric pattern not present for more symmetric loads.
Introduction
Eliminating the
traditional RF-shielded cabin (Faraday cage) used in almost all clinical MRI
installations would significantly lower scanner installation costs and simplify
siting in diverse locations such as emergency departments and intensive care
units. The Faraday cage serves two purposes. First, it attenuates
electromagnetic interference detected by the receive system which results in image
artifacts (the “receive problem”)1. Second, it attenuates out-going
electromagnetic radiation from the body Tx coil, which might interfere with
other hospital equipment (the “transmit problem”)2. The receive problem can be addressed through
external interference detection coils and post-processing algorithms to remove
detected interference from the image3-8.
The transmit problem is less well studied and is the focus of this work.
Regulatory limits restrict the peak |E|-field on a 10m radius sphere to less
than 1mV/m (= 60dBµV/m) for operational frequencies >30MHz. For proton
frequencies between 20 and 30MHz (such as 0.5T MRI), the limit is expressed as
a maximum magnetic field amplitude |B|<3.35pT (= 8.5dBµA/m) on the
10m sphere9.
In this study, we
used EM simulations to model the far-field emission levels and patterns for a
conventional CP birdcage body coil inside a solenoidal superconducting MRI
system without a shielded room. We modeled 0.5T, 1.5T, 3T, and 7T systems with
loads ranging from a uniform sphere (most symmetric) to a body model (least
symmetric).Methods
Figure 1 shows the simulated system consisting of the magnet,
cylindrical RF shield, and 32-rung high-pass CP-driven birdcage coil. Figure 1C
shows the five loads studied (arranged from the most symmetric to least
symmetric): sphere, cylinder, elliptic cylinder, symmetrized body, and body. EM
simulations were performed with ANSYS
Electronics (ANSYS Inc., Canonsburg, PA) using iterative adaptive
meshing. The mesh size was restricted to <4mm on the birdcage coil, <1mm
on lumped ports, and <20mm on the RF shield and <10mm within the load
volume. The maximum mesh size on the magnet and radiation box surfaces were
<40mm and <250mm, respectively. For each load, the birdcage coil was
tuned and matched to |S|<-20dB in order to remove imperfect
tuning/matching as a source of variability of the results. We computed the far-field
directivity, D, defined as the ratio of the radiation
intensity (= Poynting vector amplitude) in a given direction to the averaged radiation
intensity. The |E|-field pattern on the surface of the 10m radius
sphere was calculated for a CW input waveform of either: (a) 1-volt RMS total
input voltage or (b) a voltage value leading to a global SAR of 2W/kg in the load.Results
Figure
2 shows the far-field directivity and Figure 3 the |E|-field
patterns at 1.5T for the different loads. The body model produces the most
anisotropic radiation pattern at 10m distance (D is 65% greater than in the
sphere) and the greatest radiated |E|-field levels (75% (5dBµV/m)
greater |E|-field than in the
sphere).
Figure 4 shows |E|-field
patterns at 10m distance for 0.5T, 1.5T, 3T, and 7T MRI systems with sphere or
body loads (CP mode drive, 1-volt RMS input voltage). With increasing field
strength, the asymmetry and the |E|-field
level increase. For example, the 7T system has radiation pattern 3 times more
directive and generated 393-fold greater |E|
than the 0.5T MRI system. The
asymmetry pattern also undergoes a qualitative change between 3T and 7T, likely
when the bore supports waveguide propagation of a traveling wave.
Figures
5A-B show the fraction of the total input power radiated in free space, as well as the maximum |E|-field for
the different loads on the 10m radius sphere (1.5T, CP mode). Figures 5C-D show
the normalized
radiated power and maximum |E|-field for the 0.5T, 1.5T, 3T, and 7T MRI systems loaded
with the sphere and realistic body model. At constant global SAR,
the body-loaded 1.5T MRI system radiated 1.9 times more power and generated 5.1dBµV/m
more |E|-field
at 10m distance compared to the other loads. Increasing B0
from 0.5T to 7T increased the body-load radiated power 4.6×104
fold and the |E|-field
by 43.6dBµV/m. For the spherical load, these ratios were 4.1×102
and 35.8dBµV/m.Conclusion and Discussion
MRI scanners, as group II, class A devices, must adhere to
IEC 60601-1-210, which in turn refers to CISPR
119 for regulatory control of electromagnetic
radiations. Controls are specified in terms of B-field for frequencies
below 30MHz, and maximum E-field amplitudes at higher frequencies. This
requires that radiation at a 10m distance should not exceed |H|= 8.5dBµA/m
(|B|= 3.35pT) for 0.5T, and |E|= 50dBµV/m at 1.5T, and 60dBµV/m
at 3T and 7T.
The radiation patterns emerge from the sides of the magnet
cryostat and develop a strong azimuthal asymmetry with a body load. Our simulations show that the maximum
radiation by body-loaded systems without a Faraday room are 46.4dBµA/m (= 263pT),
112.6dBµV/m, 124.3dBµV/m, and 141.2dBµV/m at 0.5T, 1.5T, 3T, and 7T for a CW input
and a normalized RF exposure resulting in a global SAR of 2W/kg; which is well
over the regulatory limit. This over-radiation will be even greater in realistic MRI sequences/protocols. This indicates that, without a Faraday room, MRI radiation
limits are likely much more limiting than SAR limits.
Acknowledgements
No acknowledgement found.References
1. Harberts DW, Helvoort MV.
Sensitivity of a 1.5-T MRI system for electromagnetic fields. 2014
International Symposium on Electromagnetic Compatibility 2014:856-859.
2. Harberts DW, Helvoort
MV. Shielding requirements of a 3T MRI examination room to limit radiated
emission. 2013 International Symposium on Electromagnetic Compatibility
2013:1053-1057.
3. Rearick T, Charvat
GL, Rosen MS, et al. Noise suppression methods and apparatus (US Patent
No. 9,797,971 B2). U.S. Patent and
Trademark Office. 2017.
4. Huang X, Dong H, Qiu
Y, et al. Adaptive suppression of power line interference in ultra-low field
magnetic resonance imaging in an unshielded environment. J Magn Reson.
2018;286:52-59.
5. O'Reilly T, Teeuwisse
WM, de Gans D, et al. In vivo 3D brain and extremity MRI at 50 mT using a
permanent magnet Halbach array. Magn Reson Med. 2021;85(1):495-505.
6. Liu Y, Leong ATL,
Zhao Y, et al. A low-cost and shielding-free ultra-low-field brain MRI scanner.
Nat Commun. 2021;12(1):7238.
7. Srinivas SA, Cauley
SF, Stockmann JP, et al. External Dynamic InTerference Estimation and Removal
(EDITER) for low field MRI. Magn Reson Med. 2022;87(2):614-628.
8. Zhao Y, Xiao L, Liu
Y, et al. Predict and Eliminate EMI Signals for RF Shielding-Free MRI via
Simultaneous Sensing and Deep Learning. 2022 Asia-Pacific International
Symposium on Electromagnetic Compatibility (APEMC) 2022:213-215.
9. CISPR 11, Industrial,
scientific and medical equipment – Radio-frequency disturbance characteristics
- Limits and methods of measurement.
2016.
10. IEC 60601-1-2, Medical electrical equipment - Part
1-2: General requirements for basic safety and essential performance –
Collateral standard: Electromagnetic disturbances - Requirements and
tests. 2014.