Introduction

Potassium(³⁹K) plays a vital role in many cellular processes in neurological activities¹ as well as skeletal muscle function² and in oncology³. However, potassium imaging faces great challenges due to exceptionally low SNR arising from the low gyromagnetic ratio (~22 times lower than proton) and rapid T₂ relaxation. Clinical feasibility of ³⁹K quantification has recently been demonstrated with a commercial coil in skeletal muscle⁴. In this study, we designed and built a birdcage resonator for potassium integrating with a previously presented proton(¹H)/Sodium(²³Na) dipole/loop array⁵, with the eventual aim of imaging all three nuclei in one efficient examination. Here we present first results for the birdcage at ³⁹K.

Methods

A 15 cm diameter spherical phantom containing 300mM KCl (3M KCl, Supelco mixed with tap water) was used to characterise the coil.
Conductor comparison Prior to the birdcage construction, three loops of common conductors were built: copper strip (width=25mm, thickness=0.035mm,RS-PRO 542-5303), copper wire(diameter=1.22mm,RS-PRO 335-057), and copper tube (d=7.9mm) made by wrapping the copper tape around a hollow plastic tube. Four capacitors were distributed symmetrically along the loop coil to tune and match to 13.9MHz for ³⁹K at 7T(Fig.1). Q factors were recorded on the bench from a vector network analyser (VNA) (E5063A ENA, Keysight) when the coil was either unloaded or loaded with the phantom (Fig.2).
Birdcage construction A low-pass quadrature birdcage is designed and constructed with copper tape for potassium MRI at 7T. The resonator is composed of two end rings(d=25cm,w=2.5cm) and eight rungs(l=25cm,w=2.5cm) distributed equally around an acrylic cylinder that sat on custom-designed 3D printed holders(PLA,Ultimaker). The birdcage targets operation in two configurations, either performing on its own(Fig.3a) or inserting into an in-house developed ¹H/²³Na dipole/loop array(5) to allow image acquisition with three nuclei (Fig.3b). An analytical solution for the birdcage was derived with BirdcageBuilder⁶ and series capacitors of 379pF are inserted in the middle of the legs based on the calculation. A double probe was inserted in the middle transverse plane of the birdcage to ensure circularly polarised mode is tuned to the desired frequency. Two matching circuits made of a parallel and a series capacitor were then inserted into the sources that were 90 degrees apart on the legs to ensure a quadrature drive. Two capacitors are inserted symmetrically on both end rings to reduce eddy currents in the acquisition (Fig.3c). Cable traps tuned to the designated frequency are integrated along the coaxial cable feed (K_02252_D,Huber+Suhner, Switzerland) to inhibit common mode interference. A frequency shift of 0.27MHz was observed when the coil is placed at the centre of the bore, hence the birdcage was tuned to 13.61MHz on the bench to accommodate the shift and ensure maximum power efficiency in the acquisition.
Bench measurement S-parameters when loaded by the phantom and unloaded were recorded from the VNA and optimised tuning and matching to 13.61MHz (shifted Larmor frequency within the bore) was achieved prior to MR acquisition (Fig.4).
MR acquisitions MR images were acquired on a 7T MR scanner (MAGNETOM Terra, Siemens Healthineers AG, Germany) with the phantom. The coil was connected to a dedicated quadrature ³⁹K TR switch (Fig.3a). A 3D ultrashort echo-time sequence (TR/TE=100ms/0.27ms, FA=90deg, 1 average, 8.6×8.6×8.6mm3, 58 slices, bandwidth=30Hz/pixel, 4000 radial spokes,TA=6min12s) was used to acquire the image. No post-processing steps were applied after the acquisition.(Fig.5)

Results and Discussion

The conductor comparison showed that copper strip and tube loops exhibited distinct Q ratios despite having identical geometry, which confirms the conductor's shape influences the Q factors, as noted in previous studies⁷.(Fig.2) Copper strip produced the highest Q factor with more than double of that for copper wires, therefore it was chosen to construct the birdcage resonator. Exceptional tuning and matching was demonstrated on the bench with less than -22dB in both ports.(Fig.4) Given the narrow 0.1MHz span of the peak at such low frequency, it is crucial to recognise the frequency shift when the resonator is transported from the bench to the scanner environment and make necessary adjustments during construction. Homogeneous. phantom image were acquired SNR of 11.7(mean signal/std noise=34.8/2.9), demonstrating great potential for in-vivo 39K quantification.(Fig.5) A phantom with double physiological concentration was used for this initial experiment, further assessment of the coil will be performed with phantom of sensible concentration.

Conclusion

This study has demonstrated the feasibility of constructing a highly efficient birdcage resonator in-house for potassium MRI at 7T. Homogeneous images were acquired in a clinically feasible timeframe, underpinning the resonator's potential for in-vivo ³⁹K quantification. The integrated configuration will be explored to enable simultaneous acquisition with ¹H/²³Na/³⁹K in the future.

Acknowledgements

This work was supported by King’s China Scholarship Council, by core funding from the Wellcome/EPSRC Centre for Medical Engineering [WT203148/Z/16/Z], Wellcome Trust Collaboration in Science grant [WT201526/Z/16/Z] and by the National Institute for Health and Care Research (NIHR) Clinical Research Facility based at Guy’s and St Thomas’ NHS Foundation Trust and King’s College London. JMW and GCE acknowledge funding from the NIHR Biomedical Research Centre at The Royal Marsden NHS Foundation Trust and The Institute of Cancer Research, London, and the Royal Marsden Cancer Charity. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care.