Introduction

MR hyperpolarized ¹²⁹Xe imaging has the potential to become a mainstream imaging technique for pulmonary disease and treatment assessment. Dedicated imaging coils with a homogeneous excitation volume over the lung region are required for exceptional xenon image quality. This is a challenge for birdcage coils not centered in the RF shield. Previously conducted electromagnetic B^₁+ homogeneity analysis of multiple birdcage designs¹ identified an optimal asymmetrical xenon transmit/receive birdcage coil which was constructed for a Philips 3T Ingenia system. Imaging experiments to assess coil homogeneity were performed with a pressurized xenon gas Boltzmann cylinder and with a 6L bag containing a mixture of hyperpolarized ¹²⁹Xe gas and nitrogen placed inside a hollow custom loader phantom².

Methods

The birdcage coil was designed for a Philips 3T Ingenia 70cm bore system as a patient table mounted body coil with a detachable top half. The shell was constructed from four 3D printed endring sections, two middle sections, and two patient support inserts printed in ULTEM^TM 9085 (Stratasys). Rungs were covered by 5/8” G10 tubing mounted between the individual ULTEM^TM birdcage sections at rung locations. Shell thickness was limited to 1” along the ellipse dimensions to maximize patient space, Figure 1. Electrically, a high-pass birdcage design was implemented using the system’s proton body coil RF shield. Endrings were constructed from 1” wide and 1/8” thick copper bars and the legs from 1/4” copper tubing. Eleven rungs were distributed on the detachable top half and five rungs were distributed on the posterior table section of the coil. Electrical and mechanical connections of the birdcage's top and bottom sections were made in the lateral direction, versus traditional vertical direction, with Amphenol silver plated copper beryllium contacts. Xenon and proton baluns were placed on the I and Q drive ports to suppress common mode currents. Both ports were connected to a custom T/R switch, which connected to the MR system. Network measurements were performed with a Rohde & Schwarz ZNC3 network analyzer to assess tuning, matching, and coil Q for two adult males (84 kg and 65 kg). HFSS (ANSYS) was used for electromagnetic simulation of the asymmetrical birdcage loaded with a hollow torso loader phantom² to visualize calculated B₁⁺ distribution at different slice locations and for B₁⁺ mean deviation inside a centered 25cm x 25cm x 17cm volume. Imaging experiments were performed with a hollow torso loader phantom containing a xenon high‐pressure high‐density polyethylene vessel cylinder filled with 61% Xe and 39% O₂ at 11.6 bar² and with a 6L bag containing a hyperpolarized ¹²⁹Xe and nitrogen gas mixture. A gradient echo sequence (TR:750ms, TE:6.13ms, flip angle 74°, FOV:440 x 440mm, NSA:8, slice thickness: 100mm, recon voxel size 6.88mm x 6.88mm) was used for image acquisition of a single slice in the axial and coronal plane. An axial hyperpolarized ¹²⁹Xe image was acquired with a gradient echo sequence (TR:9.5ms, TE:2.8ms, flip angle 9.2°, FOV:300 x 221.5mm, NSA:1, slice thickness: 15mm, recon voxel size 0.94mm x 0.94mm).

Results

Loaded I and Q mode isolation was -44dB on a 84 kg male volunteer with a matching of S₁₁ at -18dB and S₂₂ at -13dB. A 65 kg male volunteer measured ‑23dB isolation, a matching of -23dB for S₁₁, and -12dB for S₂₂. Unloaded-to-loaded Q ratios for I and Q were greater than 2 for both volunteers. Electromagnetic simulation for the custom loader phantom showed a homogeneous B₁⁺ distribution in all three planes (Figure 3). Analysis of the of simulated B₁⁺ field distribution inside a centered rectangular box (25cm length, 25cm width, and 17cm height) showed 95% of the field was within 10% deviation of the mean. MR Images of the thermally polarized xenon Boltzmann phantom confirmed a homogeneous imaging profile, Figure 4. Imaging performed on a 6L bag containing a hyperpolarized ¹²⁹Xe and nitrogen gas mixture further demonstrated the homogeneous imaging performance of the coil (Figure 5).

Discussion and Conclusion

The detachable top half of the birdcage design allowed for obstruction free phantom and volunteer positioning. Connector technology and orientation eliminated the requirement for a dedicated locking feature of the two birdcage halves. The asymmetric design maximizes the available space for patients by conforming to the geometry of the MR patient table and bore wall. Mode isolation was similar to previously reported isolation of an elliptical birdcage for hyperpolarized ³He³ and better than reported for an asymmetric birdcage transmit coil for hyperpolarized ¹²⁹Xe⁴. S₁₁ and S₂₂ were close to the values found for an asymmetric birdcage transmit coil for hyperpolarized ¹²⁹Xe⁴. Electromagnetic simulation results of homogeneous excitation were confirmed with MR images on the xenon phantom and the hyperpolarized ¹²⁹Xe gas bag image.

Acknowledgements

No acknowledgement found.