Purpose

To simplify the design of a double-tuned (DT) human head 9.4T array coil while preserving relatively high transmit (Tx) performance at both frequencies.

Introduction

X-nuclei (¹³C, ³¹P etc) Magnetic Resonance Spectroscopy (MRS) and Imaging (MRSI) provide valuable tools for medical diagnostics and biomedical research. However, these methods suffer from poor SNR. Thus, increased SNR at ultra-high field (UHF, > 7T) can improve the quality of X-nuclei MRS and MRSI. In addition, these methods require use of DT RF-coils resonating at two frequencies. Commonly, UHF RF-coils must provide both local transmission and reception. This complicates the DT-design, which often consists of several layers of Tx and receive (Rx) elements. Therefore, simplification of the UHF DT-coil design without compromising its performance is very important. Good performance of the DT-array at ¹H-frequency is also important for many applications. Recently, a new type of RF coil, i.e. a dipole antenna (1), which has a simple mechanical and electrical structure, was developed and used for human body (1-3) and head (4-6) imaging at UHF. Thus, combining X-nuclei surface loops and ¹H-dipoles can simplify the design of a DT UHF human head coil. In this work, we developed and constructed a novel ¹³C/¹H human head 9.4T array coil.

Methods

Presence of ¹³C loops resonating at lower frequency (100 MHz) can still strongly affect the performance of ¹H dipole array (7). The common way of minimizing this interaction, is an introduction one or several resonant ¹H-traps (400 MHz) into each X-nuclei loop (8). Therefore, we numerically investigated how many ¹H-traps need to be introduced into each ¹³C-loop. The CST model of the DT array coil (Fig.1) included all sixteen transmit/receive (TxRx) elements, i.e. eight ¹³C loops and eight ¹H folded-end dipoles all placed in a single-layer at the same distance to the sample. The array measured 200 mm in width (left-right), 230 mm in height (anterior-posterior) and 190mm in length. The length of the ¹³C-loops measured 175 mm. Adjacent ¹³C-loops were decoupled by overlapping, and the nearest non-adjacent loops (e.g. 1&3, 2&4 etc) using transformer decoupling. Decoupling of adjacent dipole elements was provided by the RF shield (7). To improve the RF-field at the superior location (9), we added a flat RF-shield (Fig.1). We also numerically compared the performance of the DT 16-element array with that of single-tuned (ST) 8-element arrays at both frequencies. Electromagnetic (EM) simulations were performed using CST Studio Suite 2019 (CST, Darmstadt, Germany) and the time-domain solver based on the finite-integration technique. All data were acquired on a Siemens Magnetom 9.4T human imaging system. We compared the new array performance at 400MHz with that of the DT 20-element ³¹P/¹H array (8) and ST 16-loop array (10) constructed previously.

Results and Discussion:

Fig.2 shows the numerical comparison of the ¹³C/¹H array Tx-performance with that of the ST ¹H 8-element dipole array. The DT-array was simulated in two versions, i.e. ¹³C-loops having two or four ¹H-traps. Presence of ¹³C-loops decreases <B₁⁺> (averaged over 130-mm transversal slab) by ~10% (4 traps) and ~15% (2 traps). In addition, the B₁⁺ field map in the presence of 2-trap loops was stronger altered especially near the nose (Fig.3C) due to a current induced in ¹³C-loops. This current also increases SAR (Fig.3F) in this location. An addition of ¹H-dipoles had a minimal effect on the ¹³C-array Tx-performance. Averaged B₁⁺ was reduced by less than 1% and peak SAR changed only by 2%. Figs.3 shows 8x8 S₁₂ matrices obtained for the new DT-array loaded by the head-and-shoulder (HS) phantom and a human head. 400-MHz S₁₂ matrix (Fig.3A) shows the strongest coupling between adjacent dipole elements with the average value of -15.6 dB. Averaged coupling between adjacent elements at 100MHz (Fig.3B) measured -14.5 dB. The worst coupling was measured for pairs of loops separated by two elements (i.e. 1&4, 2&5 etc) with the average value of -13.1 dB. Fig.4 shows experimentally measured ¹³C B₁⁺ map. Averaged over the 130-mm transversal slab <B₁⁺> measured 32 μT/√kW (normalized to the coil input). Simulated <B₁⁺> value was 16% higher. Finally, Fig.5 shows a comparison of new DT-array with other DT- and ST-coils. The presented novel design further improves the brain coverage at ¹H-frequency without increasing the number of elements and provides the longitudinal coverage comparable with that of the substantially more complex 2x8 ¹H ST-array.

Conclusion

We showed that coupling between loops and dipoles can be minimized by placing four ¹H-traps into each ¹³C loop. The presented RF DT-array substantially simplifies the DT head coil design. At the same time, the coil demonstrated the improved ¹H longitudinal coverage and reasonable Tx-efficiency.

Acknowledgements

Funding by the ERC Starting grant / SYNAPLAST / 679927 and the Cancer Prevention and Research Institute of Texas (CPRIT) grant / RR180056 is kindly acknowledged.

References

1) Raaijmakers AJE, Ipek O, Klomp DWJ et al. Magn Reson Med 2011;66:1488-1497. 2) Oezerdem C., Winter L., Graessl A. et al. Magn Reson Med 2016;75:2553–2565. 3) Raaijmakers AJE, Italiaander M, Voogt IJ et al. Magn Reson Med 2016;75:1366–1374. 4) Clément J, Gruetter R, and Ipek Ö. Magn Reson Med 2019;81:1447–1458. 5) Connell IRO, Mennon RS IEEE Trans Med Imag 2019;38: 2177-2187. 6) Avdievich NI, Solomakha G, Ruhm L, Scheffler K, Henning A. NMR in BioMed 2020;33:1-11. 7) Shajan, G, Mirkes C, Buckenmaier K, Hoffmann J, Pohmann R, Scheffler K. Magn Reson Med 2016;75:906-916. 8) Avdievich NI, Ruhm L, Dorst J, Scheffler K, Korzowski A, Henning A. Magn Reson Med 2020;84:1076-1079. 9) Alecci M, Collins CM, James Wilson J, Liu W, Smith MB, Jezzard P. Magn Reson Med 2003;49:363-370. 10) Avdievich NI, Giapitzakis IA, Pfrommer A, Henning A. NMR in BioMed 2018, 31(9):1-13.