0911

Accelerating MRI with a Wireless Insert Gradient Coil
Edwin Versteeg1, Rik Weersink1,2,3, Sven Nouwens3, Thomas Roos1, Jeroen Siero1, and Dennis Klomp1
1Department of Radiology, UMC Utrecht, Utrecht, Netherlands, 2Department of Biomedical Engineering, Technical University Eindhoven, Eindhoven, Netherlands, 3Department of Mechanical Engineering, Technical University of Eindhoven, Utrecht, Netherlands

Synopsis

Keywords: Gradients, Gradients

Motivation: High performance gradients enable fast and high-resolution imaging but are costly and cumbersome to install in an existing MRI-system

Goal(s): Present a wireless (resonant) gradient coil setup that yields additional gradient performance without needing a dedicated amplifier

Approach: The increased gradient performance was measured using field camera measurements and acceleration performance was assessed retrospectively using phantom experiments

Results: The wireless gradient yielded a factor 23 improvement in slew rate (from 125 T/m/s to 2900 T/m/s) and 28-fold retrospective acceleration resulted in aliasing free images.

Impact: A wireless insert gradient coil enables 28-fold accelerated scanning without a supplementary gradient amplifier. This provides a cost-effective pathway for improving gradient performance with minimal system modifications.

Introduction

Insert gradient coils can be used to speed up MRI by enabling faster switching and higher gradient amplitudes without inducing peripheral nerve stimulation.1 Generally, these insert gradient coil setups require major modifications to the system and a dedicated gradient amplifier, making an insert gradient coil expensive and complex to add to an existing MRI scanner. In this work, we introduce a wireless insert gradient coil for brain imaging that does not require a dedicated amplifier and is powered by inductive coupling with the existing whole-body gradient coils. The main aim of this coil is to accelerate scans by providing an additional oscillating spatial encoding field in the phase-encoding direction like WAVE-CAIPI and bunched-phase encoding.1-3 We will quantify the performance of the wireless insert gradient coil through field camera measurements and present the first images.

Methods

An insert gradient coil can be powered wirelessly through the inductive coupling with the whole-body gradient coil. The insert gradient coil is positioned in the MR-scanner bore and the whole-body gradient field amplitude follows a sinewave trajectory. The change of magnetic flux enclosed by the insert coil loops induces a current, which can be amplified by a resonant circuit in the insert gradient coil, resulting in enlarged gradient field strengths and slew rates.
In this work, a lightweight single-axis (z-direction) insert gradient coil (Futura Composites) was used, positioned in a 7T MR-scanner (Philips Healthcare).1 The insert gradient coil has an inductance of 110 μH and was connected to a 9 μF capacitor in series to form a resonant circuit (Figure 1). The whole-body gradient coil was driven at a constant frequency of 5.25 kHz with a slew rate of 125 T/m/s, exciting the insert gradient coil at its resonant frequency. For transmit and receive, the insert gradient coil with an integrated birdcage coil was used, as well as a 32-channel receive array.

Field camera measurements (Skope) were performed to quantify the attainable amplification factor. The gradient field amplified by the insert gradient coil was compared to gradient fields without the insert gradient coil.
A phantom was imaged using a 2D gradient-echo scan with the following sequence parameters: FOV = 256 x 256 mm2, voxels size = 1 x 1 mm2, slice-thickness = 4 mm, flip-angle = 10°, echo-time = 11 ms, repetition time = 100 ms. This repetition time was used to limit the duty cycle through the capacitors. An MR-based field mapping sequence was used to estimate the gradient trajectory produced by the wireless gradient.2 Figure 2 shows the sequence schematics.
To assess the acceleration potential, the data was retrospectively undersampled. Reconstruction was performed by non-uniform fast Fourier transform (NUFFT) with density compensation and CG-SENSE implementation.1

Results and Discussion

The field camera measurements showed that the insert gradient coil amplified the gradient field 23 times at its resonant frequency (Figure 3). As a result, the slew rate of 125 T/m/s was amplified to 2900 T/m/s with a corresponding peak gradient of 50 mT/m at 5.25 kHz. The measurements show a build-up phase of roughly 8 ms before reaching a steady state. The field mapping sequence (Figure 4) showed an amplification of 2600 T/m/s, which is slightly lower than the field camera measurements. Presumably, the deviation is caused by the positioning of the insert coil and the short acquisition time (<8 ms). Figure 4 shows a phantom image acquired with the wireless insert gradient coil, undersampled by a factor of 5. More undersampling resulted in aliasing artefacts and low SNR. The slight aliasing effects in the current image are expected to originate from sparsity due to sampling in the build-up phase in resonance. This residual aliasing can be resolved using a multi-channel receive array which also allows the undersampling to be pushed to 28-fold (Figure 5). The wireless gradient approach presented in this work could yield higher acceleration factors if higher slew rates are used in combination with a multi-coil receive array. An increase in slew rate is possible by optimizing the coil design to achieve maximum inductive coupling. We do not expect PNS to be limiting, but this requires verification.1 The wireless insert coil shows great potential as a cost-effective, flexible, and easy-to-integrate solution for MRI research.

Conclusion

Imaging efficiency can be increased at least 28-fold using an insert gradient coil powered wirelessly by the whole-body gradient assembly, thereby omitting the need for expensive amplifiers.

Acknowledgements

No acknowledgement found.

References

1. Versteeg, E., Klomp, D. W. J., & Siero, J. C. W. (2022). A silent gradient axis for soundless spatial encoding to enable fast and quiet brain imaging. Magnetic Resonance in Medicine, 87(2), 1062–1073. https://doi.org/10.1002/mrm.29010

2. Bilgic, B., Gagoski, B. A., Cauley, S. F., Fan, A. P., Polimeni, J. R., Grant, P. E., Wald, L. L., & Setsompop, K. (2015). Wave-CAIPI for Highly Accelerated 3D Imaging. Magnetic Resonance in Medicine, 73, 2152–2162. https://doi.org/10.1002/mrm.25347

3. Moriguchi, H., & Duerk, J. L. (2006). Bunched Phase Encoding ( BPE ): A New Fast Data Acquisition Method in MRI. Magnetic Resonance in Medicine, 55, 633–648. https://doi.org/10.1002/mrm.20819

Figures

Figure 1. The wireless insert gradient coil (Futura) connected to capacitors (encircled) to form a resonant circuit.

Figure 2. Sequence for imaging with the wireless insert gradient coil. The sequence of Ginsert is directly induced by Gphase. The k-space sinusoidal sampling trajectory of this method is compared to a Cartesian filling.

Figure 3. 23-fold amplification of the whole-body gradient field by a wireless insert gradient coil, excited at its resonant frequency at 5.25kHz, reaching a slew of 2900 T/m/s and a peak of 50 mT/m.

Figure 4. (Left) Field measured using the MR-based field mapping. (Right) Phantom image acquired with the wireless insert gradient coil and a 1-channel receive coil, undersampled 5 times.

Figure 5. (Left) K-space filling of the acquisition with 28-fold undersampling. (Right) Phantom image acquired with the wireless insert gradient coil and a 32-channel receive coil, undersampled 28 times.

Proc. Intl. Soc. Mag. Reson. Med. 32 (2024)
0911
DOI: https://doi.org/10.58530/2024/0911