Introduction

Dipole antennas (DA) are inevitable to achieve the ultimate intrinsic signal-to-noise ratio (UISNR) in ultrahigh field (UHF) MRI¹. DA arrays, which are built of high number of channels, can provide higher acceleration factors in parallel MRI and higher SNR^2,3,4,5. For this purpose, the DA geometry requires some modifications. One of the approaches is to shorten DA using high-dielectric constant ε_r medium^6,7,8. Unfortunately, the sizes of dielectric blocks (DB) used in previous investigations were rather large. Those reports have followed what Raaijmakers et al. have suggested in their landmark paper⁶, i.e. that the height of DB should be bigger than ¼ of λ. This condition seems to be true when there is direct contact between DB and the sample. However, when DB has a higher dielectric constant ε_r (e.g. 80), and a small, 5-mm air gap is present between DB and the sample, the electromagnetic field (EMF) pattern changes substantially and there is no efficient RF power delivery to the sample. In this study, we investigate transmit field (B₁⁺) produced by four different DA designs immersed in heavy water (D₂O) by conducting EMF simulations and MR phantom experiments. We also compare in simulations three types of DA arrays including one that is comprised of a miniaturized version of one of the elements.

Methods

Electromagnetic field (EMF) and specific absorption rate (SAR) simulations were performed in spherical (diameter=85mm) and cubic phantom (size=300x300x200mm³), conductivity σ=0.4 S/m, dielectric permittivity ε_r=80) using Sim4Life (ZMT AG, Switzerland). Three different single elements were investigated: RA-S, B-T⁷, SSAD⁸. DB dimensions were: RA-S (97.5x47.6x28.6mm³), B-T (155x70x25mm³), SSAD (160x70x50mm³). RA-S geometry was scaled from ⁶ according to the higher ε_r (80 instead of 36). The elements were built of acrylic glass boxes filled with D₂O. Since our goal at this stage was a qualitative comparison, acrylic glass was not included in our simulations. Phantom experiments were conducted on a 7.0 Tesla head-only system (Magnetom, Siemens Healthineers, Erlangen, Germany) and phantom images, which are related to transmit field, were obtained using SA2RAGE imaging technique⁹. The geometry of miniaturized B-T (B-T-m) was arbitrarily modified: 105x60x5mm³. RA-S, SSAD, B-T and B-T-m were used in 8-channel DA arrays that were driven in a circularly polarized (CP) mode (phase increment/element=45°).

Results

Figure 1 shows that B₁⁺ distribution in cubic and spherical phantom is substantially different between setup A (direct contact, no air gap) and B (5-mm air gap). In setup B, the magnetic field does not propagate down towards the sample and is contained within DB. As a result, a dramatic decrease in B₁⁺ (B₁⁺/√P_in) and SAR (B₁⁺/√SAR_10g) efficiency is observed (Figure 1 and 2). Phantom measurements are in rough agreement with simulations (experimental conditions were slightly different). The observed effect is associated with the height of DB. The resulting EM field was also related to the phantom's geometry. Only element B-T (shortest distance between the antenna feed and the bottom of DB) performs well despite the air gap. In setup A, B₁⁺ efficiency is not proportional to the DB height, but rather finds its optimum around 12.5 cm (Figure 3). However, even higher DB (height=22.5cm) can produce efficient B₁⁺. Scattering parameter matrix (Figure 4) was analyzed for setup A and B (DB height=20cm). It indicates that such high DB has at least several dielectric modes in setup B. Miniaturized version of B-T in an 8-channel configuration, provides higher B₁⁺ efficiency in superficial region and only 3% lower B₁⁺ efficiency in the center of the phantom than 8-channel B-T array.

Discussion and Conclusion

We showed, that performance of DA, which are shortened using too high DB, can be completely degraded when a small air gap is present between DB and the phantom. This effect disappears when the distance between the feed of DA and the bottom of DB is reduced. Our observations disagree with what was reported previously, i.e. that the height should be at least ¼ λ. This condition remains true when there is direct contact between the block and the sample. Direct contact between DB and the sample leads to smooth EM wave propagation, and an optimal DB height to maximize B₁⁺ efficiency at given depth can be found. This is not the case when there is a sufficiently small air gap between DB and the phantom. We conclude that this effect can be related to a coupled dielectric mode of DB. This hypothesis is yet to be fully confirmed in future experiments. Our work has important implications for DA design using dielectric materials especially for brain UHF-MRI. It is rarely the case that direct contact between DB and volunteer's head can be achieved. We also extended our work, and showed in simulations, that significant reductions in overall size of DB are possible without substantial B₁⁺ efficiency decrease in 8-channel array configuration. This is very promising for high-channel count, receive-only dipole or loop/dipole arrays. There is a broad range of available, low loss dielectric materials with higher ε_r which could be considered in future DB design optimizations, and their potential benefits for UHF-MRI should be investigated.

Acknowledgements

No acknowledgement found.

References

1. Lattanzi R, et al., Approaching ultimate intrinsic signal-to-noise ratio with loop and dipole antennas. Magn Reson Med. 2018 Mar;79(3):1789-1803. 2. Roemer PB, et al., The NMR phased array. Magn Reson Med. 1990 Nov;16(2):192-225. 3. Sodickson DK, Manning WJ, Simultaneous acquisition of spatial harmonics (SMASH): fast imaging with radiofrequency coil arrays. Magn Reson Med. 1997 Oct;38(4):591-603. 4. Pruessmann KP, et al., SENSE: sensitivity encoding for fast MRI. Magn Reson Med. 1999 Nov;42(5):952-62. 5. Griswold MA, et al., Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med. 2002 Jun;47(6):1202-10. 6. Raaijmakers AJ, et al., Design of a radiative surface coil array element at 7 T: the single-side adapted dipole antenna. Magn Reson Med. 2011 Nov;66(5):1488-97. 7. Winter L, et al., Design and evaluation of a hybrid radiofrequency applicator for magnetic resonance imaging and RF induced hyperthermia: electromagnetic field simulations up to 14.0 Tesla and proof-of-concept at 7.0 Tesla. PLoS One. 2013 Apr 22;8(4): 8. Sanchez JD, et al., Radiative MRI coil design using parasitic scatterers: MRI Yagi. IEEE Transactions on Antennas and Propagation, 2018 66(3), 1570-1575. 9. Eggenschwiler F, et al., SA2RAGE: a new sequence for fast B1+ -mapping. Magn Reson Med. 2012 Jun;67(6):1609-19.