1437

Impact of Huygens Box in 3T MRI RF Safety Assessment
Xin Huang1, Xi Lin Chen2, and Shiloh Sison1

1Abbott, Sunnyvale, CA, United States, 2Abbott, Sylmar, CA, United States

Synopsis

Huygens Box is an efficient simulation technique to reduce simulation time and storage space. This paper uses simulations to investigate the impact of Huygens box on 3T MRI RF Safety Assessment. The numerical results on ASTM phantom shows the overall Symmetric Mean Absolute Percentage Error (SMAPE) average on typical MRI RF simulation is 6.28%.

Introduction

ISO/TS 109741 requires active implantable medical devices (AIMDs) intended to be used in patients who undergo an MR scan to be designed in such a way to protect patients from foreseeable hazards. Computational human models loaded inside 3T MRI RF coils are simulated to determine the exposure E field levels. Huygens Box2 (also known as Total-Field Scattered-Field (TFSF)3) is an efficient simulation technique to reduce simulation time and storage space. This approach breaks down the simulation into two stages: a) Primary simulation of the Excitation Field and b) Secondary simulation of Huygens source domain. The impact of Huygens box on E field in MRI RF Simulations is discussed in this paper.

Methods

Sim4Life4, an FDTD-based EM simulation software, is used for our simulation study. The ASTM phantom 20095 with 10 cm depth is loaded inside a typical MRI body coil (the bore diameter is 70 cm and coil length is 60 cm). The properties of the ASTM phantom are set to human tissue average εr = 78, σ = 0.47 S/m. To simulate a 3T 2-port multi-channel (MC2) MRI system, both port (denoted as I port and Q port) are simulated separately.

Two pairs of Sim4Life simulations are simulated:

  • Huygens I and Huygens Q: the empty RF body coil excited by I or Q port are simulated separately. Then I port and Q port ASTM phantom simulations are conducted using Huygens box source settings where the empty I/Q port simulation results are used as excitations.
  • non-Huygens I and non-Huygens Q: Use the actual coil source with I/Q excitation to run I port and Q port ASTM phantom simulations. This is also known as tuned I and Q.

For each of the body coil simulations, the resonant frequency simulations are performed to make sure the coil is working at the correct resonant mode at 3T RF frequency.

For data comparison, the following E field results are compared for Huygens vs. non-Huygens simulation:

  • CW simulations: the clockwise (CW) circular polarization simulation with Q to I excitation amplitude ratio 1:1 and phase different 90 degree.
  • CCW simulations: the counter-clockwise (CCW) circular polarization simulation with Q to I excitation amplitude ratio 1:1 and phase different 270 degree.

According to ISO/TS 10974, 2 methods of normalization are applied for each configuration of the E field comparison:

  • B1+avg (B1+ average) scaling: the average B1+ magnitude in the center slice is normalized to the same level (1 µT in this comparison).
  • Whole body SAR (wbSAR) scaling: the whole-body SAR of the ASTM phantom is normalized to the same level (2 W/kg in this comparison).

The SMAPE6 (Symmetric Mean Absolute Percentage Error) values are calculated to measure the overall effect of all E fields (RMS magnitude) for all 4 configurations mentioned above.

Results

The data are extracted from Sim4Life simulations. To minimize the effect of large deviation from small values, the low E field values (lowest 10 percentile of the overall E field distribution) are omitted.

The B1+ and wbSAR scaled E field comparison result is shown in Figure 1 for CW simulations and Figure 2 for CCW simulations.

Figure 3 shows the SMAPE average and standard deviation for all 4 comparison scenarios. The mean SMAPE average is 6.28%, while the mean SMAPE standard deviation is 5.05%.

Discussion

The numerical results show a good agreement between Huygens and non-Huygens E fields at 3T. The assumption of using Huygens box is that the scattered field does not significantly interfere with the source.

Conclusion

The introduction of Huygens box simulations to 3T MRI RF safety assessment is a good representation of the actual non-Huygens simulations. The difference between Huygens and non-Huygens E fields from MRI RF simulations has an averaged SMAPE of 6.28%.

Acknowledgements

No acknowledgement found.

References

  1. ISO/TS 10974: 2018 Assessment of the safety of magnetic resonance imaging for patients with an active implantable medical device; 2018
  2. A. Christ, M. G. Douglas, J. M. Roman, et al, “Evaluation of wireless resonant power transfer systems with human electromagnetic exposure limits,” IEEE Trans. Electromagn. Compat., vol. 55, no. 2, pp. 265–274, 2012.
  3. K. R. Umashankar and A. Taflove, “A novel method to analyze electromagnetic scattering of complex objects,” IEEE Trans. Electromagn. Compat., vol. 24, no. 4, pp. 397–405, Nov. 1982.
  4. Sim4Life. https://zmt.swiss/sim4life/
  5. ASTM F2182-09, Standard Test Method for Measurement of Radio Frequency Induced Heating On or Near Passive Implants During Magnetic Resonance Imaging; 2009
  6. Lucano E, Liberti M, Mendoza GG, et al. Assessing the Electromagnetic Fields Generated By a Radiofrequency MRI Body Coil at 64 MHz: Defeaturing Versus Accuracy. IEEE Trans Biomed Eng. 2016 Aug;63(8):1591-1601.

Figures

Huygens vs. Non-Huygens E Field Comparison when the CW simulations are scaled to same B1avg levels (left) and wbSAR levels (right). The ±10% reference line is shown as the red dash line.

Huygens vs. Non-Huygens E Field Comparison when the CCW simulations are scaled to same B1+avg levels (left) and wbSAR levels (right). The ±10% reference line is shown as the red dash line.

SMAPE average and standard deviation for Huygens and non-Huygens comparison

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)
1437