1415

Safety and imaging performance of 2-channel RF shimming for fetal MRI at 3T

Filiz Yetisir¹, Esra Abaci Turk^1,2, P. Ellen Grant^1,2,3, Elfar Adalsteinsson^4,5, and Lawrence L. Wald^3,5,6
¹Fetal-Neonatal Neuroimaging and Developmental Science Center, Boston Children's Hospital, Boston, MA, United States, ²Department of Pediatrics, Harvard Medical School, Boston, MA, United States, ³Department of Radiology, Harvard Medical School, Boston, MA, United States, ⁴Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA, United States, ⁵Harvard-MIT Division of Health Science and Technology, Massachusetts Institute of Technology, Cambridge, MA, United States, ⁶Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States

Synopsis

3T MRI provides increased SNR but poses technical challenges for fetal imaging such as increased field inhomogeneity and SAR. RF shimming can address some of these challenges but also adds safety considerations. Using a diverse set of 5 numerical pregnant body models generated from patient MRI datasets, we found that 2-channel RF shimming can improve transmit field amplitude and uniformity by up to 19% and 36% respectively without increasing maternal or fetal SAR. The biggest difference in SAR and transmit field patterns was observed between the supine models and the left lateral model.

Target audience

MR physicists and clinicians interested in RF safety of 3T fetal imaging.

Introduction

3T MRI can improve the quality of fetal imaging compared to the more established 1.5T MRI due to improved SNR^1,2. However, transmit field inhomogeneities and specific absorption rate (SAR) concerns at 3T can reduce this gain and limit imaging speed³. Parallel transmission (pTx)^4–6 has the potential to mitigate transmit field inhomogeneities and reduce SAR and several 3T scanners today already utilize two channel pTx systems. However, the safety of pTx systems for fetal imaging is severely understudied. Moreover, one previous study⁷ found that two-channel RF shimming can significantly increase fetal SAR and advised restriction to circularly polarized (CP) birdcage mode for pregnant subjects. They used three pregnant body models with different gestational ages (3-, 7- and 9-month-old) all artificially generated from the same non-pregnant female body model. In this study, we examine RF safety of two channel RF shimming at 3T for a more diverse set of numerical pregnant body models with differing posture, BMI, arms’ position and gestational age generated from MRI data of actual pregnant subjects⁸.

Methods

We utilize 5 pregnant numerical body models previously generated⁸; BCH1, BCH3, BCH4, BCH5, and BCH4_woA as well as the 7-month-old model by the IT’IS Foundation⁹, called "PW_II" in short. The fetal tissues in PW_II were simplified to two tissue classes, “fetal brain” and “fetus”, similar to the BCH models. These body models were loaded into a 2-channel 32-rung high pass birdcage body coil, tuned to 123.2 MHz, as shown in Figure 1. All tissue dielectric properties were assigned according to the IT’IS database¹⁰ except “fetus”, “fetal brain” (Hand et al.¹¹), and “amniotic fluid” (Peyman et al.¹²). Electromagnetic simulations were carried out using Sim4Life (Zurich MedTech, Zurich, Switzerland). Virtual observation points (VOPs)¹³ were computed with an overestimation factor of 1% and used to estimate maternal and fetal peak 10g-averaged local SAR (pSAR10g). Maternal and fetal SAR were calculated for a range of RF shim settings where the relative amplitude and phase of the two channels was varied from 0 to 2 (steps of 0.1) and -90° to 270° (steps of 10°) respectively. CP mode corresponds to a relative amplitude/phase of 1/90°. In order to quantify imaging performance, average B₁⁺ and variation of B₁⁺ (coefficient of variation, CV) inside the fetus were calculated similarly to Murbach et al.⁷ for each model and RF shim setting. All SAR and B₁⁺ values were normalized to whole-body average SAR of 2 W/kg.

Results

Figure 2 shows the average B₁⁺ and B₁⁺ variation patterns for the specified RF shim settings for all models. Imaging performance patterns are mostly similar across models except for the one in the left lateral position. Figure 3 shows overall (maternal and fetal) and fetal pSAR10g and fetal average SAR (aveSAR) patterns. In general, overall pSAR10g and fetal SAR have roughly opposite patterns as reported in Murbach et al.⁷Overall pSAR10g and fetal pSAR10g values as well as fetal aveSAR values across all models are compared in Figure 4. CP mode fetal and maternal SAR vary significantly (by a factor of 1.9-2.3) among models as reported in Abaci-Turk et al.⁸. Maternal and fetal SAR varies by a factor of 3.5-6.2 across different RF shim settings and models. The highest average B₁⁺ and lowest B₁⁺ variation achievable with RF shimming (with or without increasing maternal and fetal SAR compared to CP mode) are described in Table 1.

Discussion

RF shimming can improve average B₁⁺ and B₁⁺ variation inside the fetus by up to 28% and 49% respectively while keeping the whole-body average SAR at 2 W/kg. Without increasing maternal or fetal SAR compared to each model’s corresponding CP mode SAR levels, RF shimming can improve average B₁⁺ and B₁⁺ variation by up to 19% and 36% respectively (Table 1). Note that for models with low CP mode SAR values, restricting SAR to below that of the CP mode is more limiting compared to models with high CP mode SAR values. Safety of RF shimming for fetal MRI should be further investigated by temperature simulations^7,14 to determine SAR levels that could be considered “safe” for all subjects. This way, SAR constraints for models with low CP mode SAR values (such as IT’IS PW_II and BCH4_woA) might be relaxed, improving the imaging performance of “safe” RF shimming. Both imaging performance and SAR patterns vary most significantly between the left lateral model and the supine models. Imaging performance patterns depend on the chosen ROI, which was the entire fetus for this study (similar to Murbach et al.⁷). In the future, we will repeat the same analysis for when the ROI is the uterus. In order to improve computational efficiency in the estimate of SAR for hundreds of shim settings and 6 different body models, we used VOP matrices computed from interpolated fields (from a nonuniform grid onto a uniform grid). This introduced up to 9% error for the CP mode SAR values. Finally, motion is an inevitable part of fetal MRI and its effect on B₁⁺ and SAR patterns will be studied in the future.

Acknowledgements

This work was supported by R01EB017337, U01HD087211, R01EB006847, P41EB015896, PS1 4000002543, Siemens MR.

References

1. Krishnamurthy U, Neelavalli J, Mody S, et al. MR imaging of the fetal brain at 1.5T and 3.0T field strengths: comparing specific absorption rate (SAR) and image quality. J Perinat Med. 2015;43(2):209-220. doi:10.1515/jpm-2014-0268

2. Victoria T, Johnson AM, Edgar JC, Zarnow DM, Vossough A, Jaramillo D. Comparison Between 1.5-T and 3-T MRI for Fetal Imaging: Is There an Advantage to Imaging With a Higher Field Strength? AJR Am J Roentgenol. 2016;206(1):195-201. doi:10.2214/AJR.14.14205

3. Soher BJ, Dale BM, Merkle EM. A review of MR physics: 3T versus 1.5T. Magn Reson Imaging Clin N Am. 2007;15(3):277-290, v. doi:10.1016/j.mric.2007.06.002

4. Katscher U, Börnert P, Leussler C, Brink JS van den. Transmit SENSE. Magn Reson Med. 2003;49(1):144-150. doi:10.1002/mrm.10353

5. Zhu Y. Parallel excitation with an array of transmit coils. Magn Reson Med. 2004;51(4):775-784. doi:10.1002/mrm.20011

6. Adriany G, Van de Moortele P-F, Wiesinger F, et al. Transmit and receive transmission line arrays for 7 Tesla parallel imaging. Magn Reson Med. 2005;53(2):434-445. doi:10.1002/mrm.20321

7. Murbach M, Neufeld E, Samaras T, et al. Pregnant women models analyzed for RF exposure and temperature increase in 3TRF shimmed birdcages. Magn Reson Med. 2017;77(5):2048-2056. doi:10.1002/mrm.26268

8. Turk EA, Yetisir F, Adalsteinsson E, et al. Individual variation in simulated fetal SAR assessed in multiple body models. Magn Reson Med. 2020;83(4):1418-1428. doi:10.1002/mrm.28006

9. Gosselin M-C, Neufeld E, Moser H, et al. Development of a new generation of high-resolution anatomical models for medical device evaluation: the Virtual Population 3.0. Phys Med Biol. 2014;59(18):5287-5303. doi:10.1088/0031-9155/59/18/5287

10. Hasgall P, Gennaro F, Baumgartner C, et al. IT’IS Database for thermal and electromagnetic parameters of biological tissues. 2018;4. doi:10.13099/VIP21000-04-0

11. Hand JW, Li Y, Thomas EL, Rutherford MA, Hajnal JV. Prediction of specific absorption rate in mother and fetus associated with MRI examinations during pregnancy. Magn Reson Med. 2006;55(4):883-893. doi:10.1002/mrm.20824

12. Peyman A, Gabriel C. Dielectric properties of rat embryo and foetus as a function of gestation. Phys Med Biol. 2012;57(8):2103-2116. doi:10.1088/0031-9155/57/8/2103

13. Eichfelder G, Gebhardt M. Local specific absorption rate control for parallel transmission by virtual observation points. Magn Reson Med. 2011;66(5):1468-1476. doi:10.1002/mrm.22927

14. Hand JW, Li Y, Hajnal JV. Numerical study of RF exposure and the resulting temperature rise in the foetus during a magnetic resonance procedure. Phys Med Biol. 2010;55(4):913-930. doi:10.1088/0031-9155/55/4/001

Figures

Figure 1: Numerical model of the 2-channel, 32-rung high pass birdcage coil (a) and the numerical pregnant body models (b) used in this study. Only the skin, uterus and fetus are shown in the body models for simplicity. GA: gestational age, w: weeks, BMI: body mass index.

Figure 2: Average amplitude and the coefficient of variation (CV) of the transmit field (B₁⁺) inside the fetus for different RF shim settings where relative amplitude and phase of the two channels is varied from 0 to 2 (vertical axis) and from -90° to 270° (horizontal axis) respectively. All shim settings are normalized to maternal whole-body average SAR of 2 W/kg. CP mode: circularly polarized birdcage mode, improved imaging performance: both average B₁⁺ and B₁⁺ variation is improved compared to CP mode, no SAR increase: maternal or fetal SAR does not increase compared to CP mode.

Figure 3: Overall (maternal and fetal) peak local SAR, fetal peak local SAR and fetal average SAR for different RF shim settings where relative amplitude and phase of the two channels is varied from 0 to 2 (vertical axis) and from -90° to 270° (horizontal axis) respectively. All shim settings are normalized to maternal whole-body average SAR of 2 W/kg. CP mode: circularly polarized birdcage mode, improved imaging performance: both average B₁⁺ and B₁⁺ variation is improved compared to CP mode, no SAR increase: maternal or fetal SAR does not increase compared to CP mode.

Figure 4: Comparison of overall (maternal and fetal) peak local SAR, fetal peak local SAR and fetal whole-body average SAR values across different body models. All shim settings include all shim settings considered in this study (ones with relative amplitude larger than 2 are not included). CV: coefficient of variation, CP mode: circularly polarized birdcage mode.

Table 1: Average B₁⁺ and B₁⁺ variation (CV: coefficient of variation) across different models in circularly polarized birdcage mode (CP mode) and when using RF shimming (RFS) with or without increasing maternal or fetal SAR compared to the CP mode. The percentage of improvement in average B₁⁺ and B₁⁺ CV when using RF shimming compared to CP mode is noted in parenthesis for each case.

Proc. Intl. Soc. Mag. Reson. Med. 29 (2021)

1415