0220

Integrated AC/DC coil and dipole Tx array for 7T MRI of the spinal cord
Nibardo Lopez Rios1, Ryan Topfer2, Alexandru Foias2, Axel Guittonneau2,3, Kyle M. Gilbert4, Ravi S. Menon4,5, Lawrence L. Wald6,7,8, Jason P. Stockmann6,7, and Julien Cohen-Adad2,9

1NeuroPoly Lab, Institute of Biomedical Engineering, Polytechnique Montreal, Montreal, QC, Canada, 2Electrical Engineering, Ecole Polytechnique de Montreal, Montreal, QC, Canada, 3École Normale Supérieure de Lyon, Lyon, France, 4Centre for Functional and Metabolic Mapping, The University of Western Ontario, London, ON, Canada, 5Department of Medical Biophysics, The University of Western Ontario, London, ON, Canada, 6Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States, 7Harvard Medical School, Boston, MA, United States, 8Harvard-MIT Division of Health Sciences and Technology, Cambridge, MA, United States, 9Functional Neuroimaging Unit, CRIUGM, University of Montreal, Montreal, QC, Canada

Synopsis

Imaging the spinal cord is a never-ending challenge, especially when venturing to ultra-high fields. We propose a novel coil design for spinal cord 7T MRI, which combines state-of-the-art transmit and receive technologies. The transmit coil is a 3-dipole array, allowing homogeneous B1 field with parallel excitation. The receive part consists of a 15-channel AC/DC coil that can achieve both high sensitivity and local B0 shimming, including real-time shimming to compensate for respiratory-induced field variations. The design can achieve 32% reduction of static field inhomogeneities and 27% reduction in temporal field variance, opening the door to exciting EPI and spectroscopy applications.

Introduction

The benefits of 7T over 3T MRI are widely known; so too are the technical challenges from increased inhomogeneity in both static (B0) and RF (B1) fields.1 These dual issues are particularly problematic for imaging the spinal cord due to its position, at depth, within the vertebral column, and its proximity to the airways and lungs, which distort B0 dynamically during breathing.2,3 To resolve these issues and reap the benefits of 7T, we introduce a novel coil array for 7T MRI of the cervical spine. The design incorporates a 3-channel dipole transmit array and a neck-shaped integrated 15-channel AC/DC coil.

Methods

The design was based on T1w scans of 22 adult subjects that were averaged after registration by aligning the posterior of the neck (Fig.1).

Tx coil

Electric dipoles improve transmit efficiency compared to loop arrays over the spinal region.5 The transmit-coil (Fig.2) consisted of three 38.1cm-long dipoles. Each dipole was inductively shortened to resonate at 297MHz when loaded with the Laura body model in CST. Tuning-matching was accomplished with a co-circuit simulation. The central dipole was located 10.2cm posterior to the centre of the cervical spine. The right and left dipoles were located on a curved surface (radius: 50.6cm) to improve efficiency.

Rx coil

The former of the AC/DC array (Fig.2b) was designed (3D CAD tool) based on the average neck shape from the 22 T1w scans. 15 circular loops were consecutively created in CST by intersecting the former with a 96mm-diameter sphere. Geometric decoupling was achieved by iteratively adjusting the position/diameter of the sphere and running an S-parameters simulation. Tuning, matching and preamplifier decoupling were modeled in the schematic view by adapting published6 circuit schematics. The sensitivity profile of the array was estimated by combining (sum-of-squares) individual B1- maps. A homogeneous sample (permittivity=73, conductivity=1.1S/m) was used in all simulations.

In vivo experiment/Shim simulation

To simulate shim performance for both static ΔB0 and respiration-induced resonance offsets (RIRO), gradient-echo field maps of six adult subjects were acquired on a Siemens-3T system with subjects positioned in the 3D-printed AC/DC coil former (Fig.2b). Using the body coil in Tx/Rx mode, fields were measured over inspired and expired breath-hold conditions while a set of bellows recorded a respiratory signal. Static ΔB0 and RIRO fields in Hz were calculated7 for each subject and scaled by the nominal ratio of the system field strengths (7/2.89). Image and CAD coordinates were aligned (Fig.3) and shim basis fields in Hz/A were computed at the image voxel positions using Biot-Savart.8 Finally, shim currents (constrained to ≤2.5A per channel) were optimized for each subject, simultaneously minimizing static ΔB0 and RIRO over a region encompassing the cervical and upper-thoracic spinal cord. For comparison, equivalent optimizations were performed using the standard spherical harmonic basis up to 2nd order.

Results and discussion

Tx coil

The simulated B1+ profile, in circularly polarized (CP) mode, showed ~50% variation from the occipital lobe to T4 (Fig.4a). The transmit efficiency (per max 10-g SAR) of the CP mode, 0.27µT/√(W/kg), was 18% higher than that of the centre dipole in isolation (0.23µT/√(W/kg)); the commensurate transmit power efficiency in the centre of the cervical spine was 55nT/V. The interplay between the transmit uniformity and SAR efficiency can be adjusted through B1+ shimming. The simulated maximum coupling between adjacent dipoles was 10dB. This topology provides a practical solution to be used with receive-only coils.

Rx coil

Geometric decoupling levels ≤-17.6dB were obtained with all loops tuned and matched to 50Ω (Fig.4b). A maximum increase of 2.4dB was observed when compared to simulations with isolated adjacent pairs. The array showed a satisfactory sensitivity in the spinal cord region, with a combined B1- field variation of 27% along a sample spinal cord (blue, Fig.4b).

Shimming

Fig.5 summarizes shimming results. Across the shim volume of interest (VOI) and across subjects, the 15-channel array, when combined with a global (0th order) correction reduced the mean static ΔB0 inhomogeneity (i.e. standard deviation) from 253±23Hz using 0th-2nd order spherical harmonics down to 171±14Hz (average improvement: 32±3%). Regarding the RIRO correction, which in Fig.5 is scaled to the standard deviation of the bellows signal as measured over 1min of normal breathing, the two shim bases perform comparably well: Compared to the original RIRO, the 0th-2nd spherical harmonic basis reduced root-mean-square (RMSE) RIRO by 29±6% vs. the 27±6% reduction using the 15-channel array and global offset.

Conclusions

The proposed design can achieve 32% reduction of static field inhomogeneities and 27% reduction in temporal field variance, opening the door to exciting EPI and spectroscopy applications at 7T.

Acknowledgements

Funded by the Canada Research Chair in Quantitative Magnetic Resonance Imaging [950-230815], the Canadian Institute of Health Research [CIHR FDN-143263], the Canada Foundation for Innovation [32454, 34824], the Fonds de Recherche du Québec - Santé [28826], the Fonds de Recherche du Québec - Nature et Technologies [2015-PR-182754], the Natural Sciences and Engineering Research Council of Canada [435897-2013], the Canada First Research Excellence Fund (IVADO and TransMedTech) and the Quebec BioImaging Network [5886].

References

1. Barry RL, Vannesjo SJ, By S, et al. Spinal cord MRI at 7T. Neuroimage. 2017; 168:437-451.

2. Verma T and Cohen-Adad J. Effect of respiration on the B0 field in the human spinal cord at 3T. Magn Reson Med. 2014;72(6):1629-36.

3. Vannesjo SJ, Miller KL, Clare S, et al. Spatiotemporal characterization of breathing-induced B0 field fluctuations in the cervical spinal cord at 7T S. Neuroimage. 2017;167:191-202.

4. De Leener B, Lévy S, Dupont SM, et al. SCT: Spinal Cord Toolbox, an open-source software for processing spinal cord MRI data. Neuroimage. 2017;145(Pt A):24-43.

5. Duan Q, Nair G, Gudino N, et al. A 7T spine array based on electric dipole transmitters. Magn Reson Med. 2015;74:1189–1197.

6. Stockmann JP, Witzel T, Keil B, et al. A 32-channel combined RF and B0 shim array for 3T brain imaging. Magn Reson Med. 2016;75(1):441-51.

7. Topfer R, Foias A, Stikov N, et al. Real-time correction of respiration-induced distortions in the human spinal cord using a 24-channel shim array. Magn Reson Med. 2018;80:935–946.

8. Lin FH. Magnetic field distributions by Biot-Savart's Law. https://maki.bme.ntu.edu.tw/tools/190-2. Accessed October 4, 2018.

Figures

Figure 1. Coil configuration. 22 scans were aligned by placing a point at C2-C3 disc level on the central sagittal slice (a, white arrow). The positions of the spinal cords (red) were detected in the axial (b), sagittal (c) and coronal (d) overlaid planes by automatically segmenting the cervical cord using the Spinal Cord Toolbox.4 Rear and left views of the Tx dipoles (yellow) and AC/DC loops (white) were superposed on the sagittal and coronal images to illustrate the final position of both arrays (c, d). The coils were designed to image from the occipital lobe to ~T4 vertebral level.

Figure 2. Design details. a) Circuit schematic of a transmit dipole and b) axonometric view of the RF elements. Two ports were used in the simulations to represent the tuning inductors (LT) used for inductive shortening while a central port represented a parallel matching capacitor (CM). A bazooka balun, found to minimize coupling to the dipole with the least impact on resonant spectra, provides a balanced drive. An active detuning circuit, activated during the receive mode, was also included. Dipoles were separated by 10cm to prevent inter-element coupling from distorting B1+ profiles, while concurrently preserving the power efficiency.

Figure 3. Shim basis fields for five coils generated over the FOV (boxed region) of the acquired field maps. Vitamin E capsules (hidden above by the coil overlay) were positioned on the posterior surface of the former at the centers of the medial coils (2, 10, 15) to provide fiducial markers by which the coil geometry (exported from CST) was matched to the DICOM images.

Figure 4. RF simulations. a) B1+ profiles of individual dipoles and combined field (CP mode). The dipole array was rotated ~4° to compensate small asymmetries in individual B1+ profiles; this allowed the outer dipoles to produce the equivalent B1+ magnitude at the centre of the cervical spine, thereby potentially limiting the range in magnitudes required for B1+ shimming and thus the required peak voltage. b) B1- profiles of the five receive loops with the main contributions to the central axial image of the cervical spine, the medial sagittal plane and two 6cm-separated axial slices. S-parameters are shown on the top-right.

Figure 5. Static ΔB0 (top) and respiration-induced resonance offsets (RIRO, bottom) under four conditions: (i) The acquired fields scaled to 7 T (Original); (ii) optimized using ideal spherical harmonics up to second order (0th-2nd SH); (iii) optimized using the 15-channel multi-coil along with a global frequency offset (0th + MC); and (iv) optimized concurrently using all terms (0th-2nd SH + MC). Optimizations were restricted to the shim volume of interest (red outline) which was similar across subjects. The left panel exhibits the results for the subject (2) of Fig.3 and the right panel summarizes results across subjects.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)
0220