Introduction

²³Na-sodium MRI keeps increasingly demonstrating diagnostic value in a multitude of studies and clinical applications due to its capability to provide information on tissue viability^[1,2]. Yet, sodium MRI suffers from weak SNR and low abundance of sodium in the human body.Therefore, ¹H MR images are needed for morphological information to compensate for the poor resolution of sodium MRI. In order to co-register ²³Na and ¹H MR images, a double resonant ²³Na/¹H RF system is the optimal solution. Since clinical scanners do not support X-nuclei hardware integrated in the MR system, local coils have to enable combined proton and X-nuclei MRI. In this work we present a clinical double-resonant RF system consisting of a shielded ²³Na BC coil^[3], a 16 channel ²³Na Rx array and a local ¹H Helmholtz coil inside the shielded ²³Na BC coil. The complete system is demonstrated in EM simulations and initial feasibility measurements are performed.

Methods

Simulation:
EM fields were calculated using FEM simulations (CST – Computer Simulation Technology GmbH) with a tetrahedral mesh. Mesh refinement converged for all simulations to at least 5% for all S-parameters at the resonance frequency 32.586 MHz (²³Na) and 123.2 MHz (¹H).
²³Na Tx:
A quadrature driven 16-legged asymmetric ²³Na was modeled following reference^[3].
²³Na Rx:
A 16 channel ²³Na Rx only array was designed with 8 posterior and 8 anterior channels to be used for the abdomen. The Rx elements were decoupled iteratively using EM simulations. Each Rx coil was split twice and a ¹H trap was added. The H-fields of the ²³Na Rx channels were combined using the matched filter approach^[4].
¹H TxRx:
The ¹H TxRx coil was constructed as a Helmholtz coil covering the FoV of the ²³Na Rx array. Each loop of the Helmholtz coil was equally split 16 times.
Phantom:
Four cylindrical phantoms were used with diameter = 115 mm, length = 205 mm, σ = 0.91 S/m and ε = 80.

The simulation setup of the different coils separately and the complete simulation setup are shown in Figure 1 a).

Measurement:
For initial feasibility measurements, the posterior part of the setup consisting of an 8 channel ²³Na Rx only coil (Figure 2) combined with one ¹H TxRx posterior coil was built using the dimensions of the EM simulation. Four cylindrical phantoms (Siemens Healthineers, dimensions and solution as described above) were used for phantom measurements.
We used a density adapted 3D radial sequence^[5] for sodium imaging (TE/TR/FA=0.54ms/50ms/54°, 8500 projections, FoV=(450mm)³, resolution=(6mm)³, T_RO=20ms), and a 2D gradient echo sequence for proton imaging (TE/TR/FA=10ms/358ms/90°, FoV=450mm², matrix=256²,Res = 1.75 mm, slice=5mm).
The reference scan was acquired using the ²³Na BC coil stand alone in TxRx mode.
In order to protect the detuning PIN diodes of the ²³Na BC coil it stayed resonant during the ¹H measurement but the ²³Na Rx array was detuned.

For SNR calculation a noise scan was performed using the same parameters as for the signal scan but without Tx power applied. The ²³Na Rx channels were combined identical to the simulation using the matched filter approach.

Results

Simulation:
The center coronal planes of the simulated B₁-maps (normalized to 1W accepted power) are shown in Figure 1 b). The ²³Na array showed a 2-fold B¹-field increase compared to the ²³Na BC coil. The ²³Na array yielded an up to 6-fold higher B¹-field compared to the ²³Na BC coil near the Rx coils.
The ¹H TxRx coil covered almost the whole FoV of the ²³Na Rx array but showed a slightly unbalanced behavior in the longitudinal direction.

Measurement:
The center coronal planes of the SNR measurements are shown in Figure 3. Consistent with the simulations, the 8 channel ²³Na Rx only array already achieved a 2-fold SNR increase in the coronal plane and an up to 6-fold SNR increase near the Rx coils compared to the ²³Na BC coil.
The ¹H SNR map revealed a similar FoV as the ²³Na Rx only array. Yet, the coil profile was unbalanced in longitudinal direction.

Discussion & Conclusion

A double resonant ¹H/²³Na whole-body RF setup was demonstrated for clinical MRI at 3T. The feasibility of a 16 channel ²³Na Rx only array combined with a ¹H Helmholtz coil inside a ²³Na BC coil was proven in simulation. A 2-6 fold improvement of the ²³Na array was achieved over the whole FoV in simulation. This performance could also be reproduced in initial SNR measurements of a reduced setup. Yet, the rudimentary ¹H coil revealed an unbalanced coil profile and low SNR. Future research should focus on getting ¹H performance near to a clinical level. The RF setup should additionally be evaluated in in-vivo scans.

Acknowledgements

No acknowledgement found.

References

[1] Madelin G, Ravinder RR. Biomedical applications of sodium MRI in vivo. J Magn Reson Im. 2013;38(3):511-529.
[2] Zöllner FG, Konstandin S, Lommen J, et al. Quantitative sodium MRI of kidney. NMR Biomed. 2016;29(2):197-205.
[3] Wetterling, F., Corteville, D.M., Kalayciyan, R. et al., Whole body sodium MRI at 3T using an asymmetric birdcage resonator and short echo time sequence: first images of a male volunteer. Phys Med Biol, 2012;57(14), p.4555.
[4] Schnell W, Renz W, Vester M, Ermert H. Ultimate signal-to-noise-ratio of surface and body antennas for magnetic resonance imaging. IEEE T Antenn Propag, 2000;48(3):418-28.
[5] Nagel AM, Laun FB, Weber MA, et al. Sodium MRI using a density‐adapted 3D radial acquisition technique. Magn Reson Med.2009;62(6):1565-1573.