Synopsis

Keywords: RF Arrays & Systems, RF Arrays & Systems

Dipole antennas can be combined with loop elements to increase a central SNR in human brain MRI at 7T (300MHz). However, this approach was not investigated at lower frequencies. In this study a degenerate birdcage coil was combined with a pair of bent dipole antennas to increase a central SNR for phosphorus MRS of human brain at 7T (120MHz). Simulations showed that this approach could provide a 1.6-fold central SNR gain vs. degenerate birdcage-only. The RF coil was constructed and successfully tested at the bench. In the next step, an experimental validation of the proposed strategy will be performed.

Introduction

Phosphorus (³¹P) MRI/MRS is a powerful tool to probe brain metabolism and physiology in vivo¹. So far most studies focused on the investigation of cortical areas. However, subcortical brain areas (e.g. basal ganglia, nucleus accumbens, thalamus, etc.) play essential roles in the regulation of a diversity of behaviors and cognition, and their dysfunction is implicated in several neurological and psychiatric disorders. Therefore, applying ³¹P MR studies in subcortical areas is of great interest. Despite using ultrahigh (UHF) magnetic field strengths, signal-to-noise-ratio (SNR) in subcortical regions (central region of the human brain) is still unsatisfactory. Unlike for peripheral regions, the SNR in the center of the brain is not expected to be improved by increasing the number of loop elements in the array². However, recent work showed that combining loop elements with dipole antennas can be used to improve SNR in the center of human brain at UHF (300 MHz)³. To the best of authors’ knowledge, so far this strategy was not investigated at 120 MHz. This can be more challenging due to the size of a dipole antenna at lower frequency and not as strong coil-sample coupling than as at 300 MHz. Notwithstanding a lacking experimental evidence, we decided to pursue this idea inspired by a theoretical model provided by Lattanzi et al.⁴ Therefore, the goal of this work was to develop a degenerate birdcage coil mixed with bent dipole antennas for ³¹P MRI/MRS of human brain at 7T which could provide a central SNR increase.

Methods

An 8-channel degenerate birdcage with 240 mm diameter for 120.3 MHz was designed similarly as reported earlier⁵. The length of each loop was 220 mm with 8 capacitors per loop. The copper width was 10 mm. Two large dipole antennas surrounding an average human head were incorporated into the birdcage design (Fig.1). Due to the significant length of the dipole antennas at 120.3 MHz (1.2m), they were bent and crossed, as demonstrated by Hong et al.⁶ This eliminated dipole-dipole coupling and enabled using smaller inductors for tuning. In our approach dipole antennas were intended to be used as receive-only elements. The RF feeds for ³¹P dipole antennas were located at the top of the head. The array was loaded with a spherical phantom (radius=85mm, ε_r=78, σ=0.4S/m). A 4-channel ¹H dipole antenna array for anatomical imaging and B₀ shimming was incorporated into the design. Each ¹H RF feed point overlapped with the center of the corresponding ³¹P loop. The ³¹P RF coil was equipped with a 1:8 power divider and 8 transmit/receive switches (MRCoilTech,UK). Simulations were performed using Sim4Life (Zurich Medtech,Switzerland). Transmit field efficiency was defined as B₁⁺/√P, where P is the input power. SNR was evaluated using an implementation of the Roemer’s algorithm⁷ which was based on the S-matrix formalism proposed earlier⁸. The same multi-channel simulation was used to investigate three scenarios: a) degenerate birdcage and dipole antennas on resonance, b) degenerate birdcage on resonance and dipole antennas open at feeding points, c) “standalone” degenerate birdcage coil (i.e. all ports in each dipole antenna defined as open making the antenna segments short). The combined ³¹P/¹H array was constructed and evaluated (tuning, matching, decoupling) at the bench (Fig.2). Bench measurements were performed when the RF coil was placed inside a dummy head gradient insert.

Results

S-matrix was obtained from simulations (Fig.3): coupling between adjacent loop elements below -15dB (loop-loop) and close to -10dB (up to -15dB) for non-adjacent ones. Coupling between ³¹P dipole antennas was below -40dB, and between dipole antennas and adjacent loop elements was close to -10dB. B₁⁺ efficiency in the center of a spherical phantom was found to be 1.70 μT/√W (when ³¹P dipole antennas were open in TX) and 1.69 μT/√W for a “standalone” degenerate birdcage (Fig.4). SNR in the center of a spherical phantom calculated for a combination of degenerate birdcage and bent dipole antennas was the highest (Fig.5): it was approximately 1.6- to 1.7-fold higher than for 8-channel degenerate birdcage alone.

Discussion

In this work a degenerate birdcage coil was for the first time combined with a pair of bent dipole antennas to increase a central SNR for phosphorus (³¹P) MR of human brain at 7T. Simulations indicate that this approach should provide a 1.6-fold gain in central SNR vs. 8-channel degenerate birdcage. It is expected that in vivo SNR gain will be slightly lower when losses from inductors, cables, etc. will be included. However, the data are already very promising and provides an encouragement to validate them experimentally. Ongoing measurements at the bench with human subjects suggest that non-adjacent loop-loop coupling is lower than in simulations. The main limitation of our study is a lack of experimental validation, yet this will be provided in the nearest future given that the RF coil was already designed and successfully tested at the bench.

Acknowledgements

The authors acknowledge access to the facilities and expertise of the CIBM Center for Biomedical Imaging, a Swiss research center of excellence founded and supported by Lausanne University Hospital (CHUV), University of Lausanne (UNIL), Ecole polytechnique fédérale de Lausanne (EPFL), University of Geneva (UNIGE) and Geneva University Hospitals (HUG).

References

1. Xin et al., Frontiers in Nutrition 2018.
2. Wiggins et al., MRM 2010.
3. Avdievich et al., MRM 2022.
4. Lattanzi et al., MRM 2018.
5. Brown et al., Neuroimage 2016.
6. Hong et al., IEEE Trans on Med Imag 2018.
7. Roemer et al., MRM 1990.
8. Kuehne et al., MRM 2015.