2658

Optimal array configuration for cerebral cortex MRI at 7T: six center-fed dipoles with two loops RF coil array

Jérémie Daniel Clément¹, Rolf Gruetter^1,2,3, and Özlem Ipek⁴

¹CIBM - LIFMET, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, ²Department of Radiology, University of Geneva, Geneva, Switzerland, ³Department of Radiology, University of Lausanne, Lausanne, Switzerland, ⁴CIBM-AIT, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland

Synopsis

Searching for optimal RF coil array for whole-brain MR applications, central-fed dipole was compared with fractionated dipole in a single and RF coil array configurations with combination of loop coils in measurements and electromagnetic field simulations in terms of B₁⁺ transmit efficiency, field uniformity, SAR and mutual coupling in an array. High transmit field performances were shown over the cerebral cortex by using phase-only shimming with six center-fed dipoles with two loops RF coil array on experimentally measured B₁⁺-maps and anatomical MR images.

Introduction

Dipole antennas were proposed as an alternative to loop coils or microstrips as they were shown to have better penetration depth and field uniformity, at 7T¹. Several earlier designs such as radiative, bow-tie, fractionated, snake or dipole-loop configurations have been investigated for body and brain magnetic resonance (MR) imaging^1-6. In this study, we made a comparison between a central-fed dipole and a fractionated dipole configuration with respect to B₁⁺ transmit efficiency, excitation uniformity, SAR and mutual coupling in a single coil and 8-channel coil array.

Methods

Center-fed and fractionated dipole array designs: An 8-channel center-fed dipole and an 8-channel fractionated dipole array were built with copper printed on flexible PCB and tuned/matched to 297.2MHz/50Ohms. B₁⁺-maps⁷ were experimentally measured on a spherical phantom for single center-fed and fractionated dipole (170mm-length,25mm-width, an inductor placed in the fractionated part) (Fig.1). S-parameter matrices of each array were measured on a network analyzer (Agilent Technologies 5071C-ENA Series,USA).

Electromagnetic Field Simulations: Finite difference time domain (FDTD) simulations were performed on Sim4Life 3.1 (ZMT,Switzerland) on a Virtual Family human model⁸ for three RF coil array configurations: an 8 fractionated dipole array; a six fractionated/six center-fed dipole array with two loop coils covering the frontal lobe region. All coils were tuned/matched to 297.2MHz/50Ohms (>-15dB).

Six center-fed dipole with two loop coil array design: Compared to placing two dipoles on the frontal lobe, two loop coils (95x85mm² each) demonstrated higher transmit field efficiency and lower mutual coupling by coil overlapping. While five of the dipoles were placed around the occipital and temporal lobes of the brain, the sixth dipole was placed around the parietal lobe to provide additional B₁⁺ in the upper region of the brain. Dipoles were copper-printed (35μm-thickness,25mm-width) on flexible PCB (FR4,0.1mm-thickness) and dipole length was varied from 157mm to 186mm according to the shape of the brain. The coils were tuned/matched (S11>-15dB,S12>-8.5dB) using fixed/variable capacitors and inductors. Common-modes on the coaxial cables were diminished with balun-transformers. Conformal housing (Fig.2) was designed (Solidworks,DassaultSystems,France) and 3D-printed (EOSINT P395,EOS,Germany) in nylon (EOS,PA2200).

MR Experiments: B₁⁺-maps and anatomical images from a healthy male volunteer were acquired using a Magnetom 7T MR scanner with an 8x1kW RF-amplifier (Step1,Siemens,Erlangen,Germany). Optimal RF phases⁹ were calculated with a PSO method¹⁰ and the following sequences were used: 3D-Turbo-Spin-Echo (0.8x0.8x1mm³,TR/TE=2000/120ms,TA=10min,GRAPPA=2,Averages=1), multi-slice GRE (0.3x0.3x3mm³,TR/TE=1000/16ms,TA=5min30s,GRAPPA=2),3D-MP2RAGE (2x2x2mm³,TR/TE =5500/1.67ms, TA=5min,GRAPPA=2).

Results

While the center-fed dipole yields higher transmit efficiency in deeper regions, the fractionated dipole provides larger longitudinal field coverage (Fig.1). Measured mutual coupling for 8-channel center-fed and fractionated dipole arrays demonstrate similar results (S12>-8.1dB and S12>-7.4dB, respectively). Moreover, the $$$B_{1,mean.ROI}^+/\sqrt{SAR_{10g,max}}$$$ value does not exhibit significant differences for center-fed and fractionated arrays (0.50 vs 0.53$$$\mu T\sqrt{kg}/W$$$). As no substantial gain is observed using the fractionated dipole, the central-fed array was further investigated. Six center-fed dipoles with two frontal loop coil array exhibits a high transmit field in close vicinity to the array element (Fig.3B). However, coupling is visible between dipoles 3 and 2 and dipole 5 with loop 2 (Fig.3A). A mean flip angle of $$$90°\pm16° (SAR_{10g,max}=0.89W/kg)$$$ and $$$90°\pm6° (SAR_{10g,max}=0.55W/kg)$$$ is achieved at 84V/channel and 96V/channel, over the ROIs, by using phase-only RF shimming on central transverse and sagittal slices (Fig.4A). High-resolution (0.3x0.3x3mm³) multi-slice GRE images (Fig.4B) demonstrate high signal uniformity. High-resolution (0.8x0.8x1mm³) 3D-TSE images (Fig.5A-B) exhibit good coverage of cerebral cortex and homogeneous signal over multiple slices in the brain. The dark line visible in the frontal lobe is an inversion band resulting from the high transmit efficiency of the two loops in their close vicinity. 3D-MP2RAGE images (Fig.5C) demonstrate that whole-brain is covered by phase-only RF shimming applied on a sagittal slice.

Discussion and conclusion

Optimal array configuration for cerebral cortex imaging was shown for a six center-fed dipole, two frontal loop coil array by B₁⁺-maps and anatomical images with phase-only RF shimming. This conformal coil array design exploited the placement of 6 dipoles on the temporal, occipital and parietal lobes of the brain, and two loop coils on the frontal lobe. No decoupling treatment was applied for the neighboring dipoles. The 8-channel array presented here demonstrates high transmit performance allowing any desired flip angle over the cerebral cortex by using phase-only RF shimming. High coverage and good field uniformity were obtained on high-resolution (0.8x0.8x1mm³) TSE, high-resolution (0.3x0.3x3mm³) GRE and MP2RAGE images. Future work will include the construction of independent receivers to increase the signal and allow for higher acceleration factors. We concluded that this 8-channel coil array configuration is remarkably efficient, in terms of coverage and B₁⁺ performances, and can be used for multiple studies in the human brain at 7T.

Acknowledgements

This study was supported by Centre d’Imagerie BioMédicale (CIBM) of the UNIL, UNIGE, HUG, CHUV, EPFL and the Leenaards and Jeantet Foundations.

References

[1] Raaijmakers AJE et al. Magn. Reson. Med. 2011; 66: 1488-1497

[2] Oezerdem C et al. Magn. Reson. Med. 2016; 75(6): 2553-2565

[3] Duan Q et al. Magn. Reson. Med. 2015; 74(4): 1189-1197

[4] Raaijmakers AJE et al. Magn. Res. Med. 2016; 75(3): 1366-1374

[5] Steensma B et al. Proc. Intl. Soc. Magn. Reson. Med. 24 (2016): 395

[6] Chen G et al. Proc. Intl. Soc. Magn. Reson. Med. 23 (2015): 3133

[7] Eggenschwiler F et al. Magn. Reson. in Med. 2012; 67: 1609-1619

[8] Gosselin MC et al. Phys. in Med. and Biol. 2014; 59(18)

[9] Metzger GJ et al. Magn. Reson. Med. 2008; 59(2): 396-409

[10] Clement J et al. Proc. ESMRMB, 503, 2016

Figures

Fig. 1 – Experimentally measured B₁⁺ maps on phantom for a single center-fed and fractionated dipole. The slice was taken along the longitudinal axis of the dipoles. Isolines are represented in red with their corresponding value inserted.

Fig. 2 – Photos of the general design of the 8-channel array in three views. Five center-fed dipoles (Dipoles 1 to 5) are placed around occipital and temporal lobes of the brain and one (Dipole 6) is placed around parietal lobe. Two loop coils (1 and 2) are placed over the frontal lobe of the brain.

Fig. 3 – A) Measured scattering matrix and B) Experimentally measured individual B₁⁺ maps for 500 μs hard pulse (90°voltage input = 167V/channel), for six center-fed dipole with two frontal loop coil array. The transverse slice was taken in the midbrain region. In each map, active coil is represented as a black line.

Fig. 4 - A) RF phase-shimmed B₁⁺ maps for 500 μs hard pulse (90° voltage input = 96V/ch. in transverse and 84V/ch. in sagittal). Phase-only RF shimming was applied in various ROIs (white dashed lines) B) High-resolution multi-slice GRE images (0.3 x 0.3 x 3 mm³) acquired in the transverse and sagittal planes with the RF phases used in A).

Fig. 5 - High-resolution 3D TSE images (0.8 x 0.8 x 1 mm³) displayed in sagittal (A) and transverse (B) planes for different slices. No post-processing correction was applied. The RF phases applied for A) and B) correspond to Fig. 4A settings. The dark line visible in the frontal lobe (A) is an inversion band resulting from the high efficiency of the loops in their close proximity. C) 3D MP2RAGE (2 x 2 x 2 mm³) image displayed in the three main orientations. RF phase shimming was applied in the sagittal plane, aiming to cover the whole-brain.

Proc. Intl. Soc. Mag. Reson. Med. 25 (2017)

2658