4652

In-vivo image acceleration with an 8-channel local B0 coil array and parallel imaging in a 9.4T human MR scanner
Rui Tian1, Theodor Steffen1, and Klaus Scheffler1,2
1High-Field MR center, Max Planck Institute for Biological Cybernetics, Tuebingen, Germany, 2Department for Biomedical Magnetic Resonance, University of Tübingen, Tuebingen, Germany

Synopsis

Keywords: New Trajectories & Spatial Encoding Methods, New Trajectories & Spatial Encoding Methods, nonlinear gradient encoding

We further developed a recent idea called spread spectrum MRI to reduce the sampling time by rapidly modulating spins with localized magnetic fields during signal readout. Given phantom experiments tested and safety evaluation for human subjects performed, this time, we started in-vivo measurements of human head with multi-slice FLASH sequence accelerated by local B0 coil modulations, and examined the reconstructed image quality. Sinusoidal modulation schemes in various phase offset patterns and frequencies given different scanner bandwidth were tested and compared, which were shown to boost the image acceleration from 6-fold (i.e., SENSE only) to about 8-fold in one phase-encoded dimension.

Introduction

Image acceleration can be achieved with various types of gradient insert, by performing phase and frequency encoding with tailored nonlinear B0 fields for improved spatial encoding efficiency (i.e., PATLOC1, O-space2), or by sinusoidally modulating a linear or nonlinear B0 field to sample more k-space information per unit time (i.e., Wave-CAIP3 with head gradient insert4, FRONSAC5), to mention but a few. Here, we further develop a recent idea called spread spectrum MRI6, which utilizes rapid modulation of localized magnetic fields produced by a set of local B0 coils and imposes unique spin phase evolution during linear gradient readout to disentangle signals between local regions and therefore, accelerate scans.

Our 8-channel local B0 array, compatible with an 18-transmit/32-receive RF coil7 for in-vivo head imaging at 9.4T, was tested to speed up MRI sampling with phantom experiments8 and passed the safety evaluation with respect to PNS and acoustic noise level for human subjects9. This time, we conducted in-vivo experiments of 2D FLASH10 scans with sinusoidal modulations of the local B0 coils and examined the reconstructed image quality from the retrospectively undersampled datasets.

Methods

The system hardware setup was similar to before8 (Figure 1). Connected with 8 power amplifier channels (IECO, Finland), the local B0 coil array (8 square loops, each 10cm x 10cm, 14 windings) was fixed in the patient bed of a 9.4T whole-body human MR scanner (Siemens Healthineers, Erlangen, Germany), with a shielded 18-transmit/32-receive RF coil inserted in its support7. To keep track of the spin phase evolution of the object imposed by the local coils with high accuracy, which is vital to image reconstruction, the magnetic field map produced by each local B0 coil was obtained and combined with the real-time current monitor of the power amplifier, following the same procedure as before8.

In-vivo multi-slice FLASH scans (here, 6 slices) in 1.1 mm(readout) x 1.1 mm(phase) x 2.5 mm(slice) were performed with the local B0 coils driven by sinusoidal currents during the signal readout, on a healthy adult volunteer in agreement with the institution’s ethics policy. The 2D FLASH scans with several sinusoidal modulation schemes were examined to test the capability of our setup for image acceleration and to compare encoding efficiency between distinct phase offset patterns (i.e., nonlinear field shapes) and frequencies, respectively. The images were reconstructed in hybrid space similar to the Wave-CAIPI3, without additional penalty terms.

The experiments were designed as below. First, two different phase offset patterns of sinusoidal currents in the 8 channels were defined (Figure 1). Specifically, the phase pattern A forms a rotated nonlinear field: $$ (0^{\circ},45^{\circ},90^{\circ},135^{\circ},180^{\circ},225^{\circ},270^{\circ},315^{\circ})$$
and, the phase pattern B forms an oscillating nonlinear gradient roughly aligned with the phase encoding direction:$$ (0^{\circ},0^{\circ},0^{\circ},180^{\circ},180^{\circ},180^{\circ},180^{\circ},0^{\circ})$$

Next, the experiments were conducted in three groups:
1. Figure 2 Low bandwidth: 200 Hz/Pixel. Local coils: 7kHz/50Apk, the phase pattern B.
2. Figure 3.: Low bandwidth: 200 Hz/Pixel. Local coils: 10kHz/50Apk, the phase patterns A and B.
3. Figure 4, 5. High bandwidth: 380 Hz/Pixel. Local coils: 7kHz, 10kHz, 12kHz, and 13kHz, all with 50Apk and the phase pattern B.

Results

Ranged from 6-fold to 10-fold acceleration along one phase-encoded dimension, the reconstructed images and the geometric (G) factor maps from modulation experiments as well as reference scans were shown in Figure 2-5. Only a single slice (i.e., #2) was shown for simplicity, although not the most strongly modulated.

In Figure 2, a comparison between the image acceleration with SENSE alone and with the additional local B0 modulation was made. Strong noise amplification was found in the 7-fold SENSE reconstruction. However, there was very little artifact in the 8-fold accelerated image with local coil modulation, and the SNR loss increased much slower than the SENSE reconstruction, especially in the peripheral regions of the objects, as the acceleration factor grew further.

In Figure 3, the phase offset patterns of sinusoidal currents (Figure 1) were compared. For the 6-fold and 7-fold image acceleration, the difference in the encoding efficiency was negligible. However, as the acceleration factor increased, phase pattern B started to become more SNR efficient. Additionally, the G map values for modulations in 10kHz given the low bandwidth are generally larger than the ones in 7kHz (Figure 2).

In Figure 4 and 5, sinusoidal modulations in higher frequencies were compared. For the acceleration factor equal to or above 8, both the reconstructed images and the G maps demonstrate the SNR advantages for the relatively lower modulation frequency. Additionally, the current peak value was reduced to 45A for the 10kHz, and 40A for the 12kHz and the 13kHz, due to the damping of the power amplifier equal to or above 10kHz.

Discussions/Conclusion

Our system has been tested successfully to substantially boost the MR sampling speed on the top of SENSE for in-vivo FLASH scans, increasing the effective acceleration factor along one dimension roughly from 6 to 8. The encoding efficiency, especially in the peripheral regions, does not decrease so rapidly as SENSE, for acceleration factor beyond 8. The various experiments in comparison illustrate the possibility to create different sinusoidal modulation schemes with the local B0 coils setup, as an invaluable degree of freedom to design novel encoding schemes in future work.

Acknowledgements

The first author thanks Mr. Oliver Holder in the EE workshop of the MPI-KYB for quickly repairing the cable malfunction in the local B0 coils, before the deadline of this abstract.

This study is supported by ERC Advanced Grant No 834940.

References

1. Hennig J, Welz AM, Schultz G, et al. Parallel imaging in non-bijective, curvilinear magnetic field gradients: a concept study. Magn Reson Mater Phys Biol Med. 2008;21(1-2):5. doi:10.1007/s10334-008-0105-7

2. Stockmann JP, Ciris PA, Galiana G, Tam L, Constable RT. O ‐space imaging: Highly efficient parallel imaging using second‐order nonlinear fields as encoding gradients with no phase encoding. Magn Reson Med. 2010;64(2):447-456. doi:10.1002/mrm.22425

3. Bilgic B, Gagoski BA, Cauley SF, et al. Wave-CAIPI for highly accelerated 3D imaging: Wave-CAIPI for Highly Accelerated 3D Imaging. Magn Reson Med. 2015;73(6):2152-2162. doi:10.1002/mrm.25347

4. Versteeg E, Klomp DWJ, Siero JCW. A silent gradient axis for soundless spatial encoding to enable fast and quiet brain imaging. Magn Reson Med. 2022;87(2):1062-1073. doi:10.1002/mrm.29010

5. Wang H, Tam LK, Constable RT, Galiana G. Fast rotary nonlinear spatial acquisition (FRONSAC) imaging. Magn Reson Med. 2016;75(3):1154-1165. doi:10.1002/mrm.25703

6. Scheffler K, Loktyushin A, Bause J, Aghaeifar A, Steffen T, Schölkopf B. Spread‐spectrum magnetic resonance imaging. Magn Reson Med. Published online April 26, 2019:mrm.27766. doi:10.1002/mrm.27766

7. Avdievich NI, Giapitzakis IA, Bause J, Shajan G, Scheffler K, Henning A. Double-row 18-loop transceive-32-loop receive tight-fit array provides for whole-brain coverage, high transmit performance, and SNR improvement near the brain center at 9.4T. Magn Reson Med. 2019;81(5):3392-3405. doi:10.1002/mrm.27602

8. Tian, R., Loktyushin, A., Buckenmaier, K., Steffen, T., Scheffler, K. (2022, May). “Image acceleration with an 8-channel local coil array compatible with parallel imaging in a 9.4T human MR scanner.” In proceedings of the joint annual meeting of ISMRM-ESMRMB 2022, London, UK.

9. Tian R, Davids M, Thielscher A, Buckenmaier K, Steffen T, Scheffler K. (2022, October). Safety evaluation of an 8‐channel local coil array with sinusoidal modulation in a 9.4T human MR scanner. ISMRM workshop on MR safety 2022, New York, USA.

10. Haase A, Frahm J, Matthaei D, Hanicke W, Merboldt KD. FLASH imaging. Rapid NMR imaging using low flip-angle pulses. Journal of Magnetic Resonance. 1986;67(2):258-266. doi:10.1016/0022-2364(86)90433-6

Figures

Figure 1. The B0 coil setup generates an additional magnetic field. Left: The local B0 coil array in a 9.4T scanner, with an RF coil with 32 receive elements inserted within its support. Right: the 50Apk sinusoidal current with phase offset patterns A and B, generating different nonlinear magnetic fields (A, E), corresponding to the gradient distributions (B, F) reaching up to 16 and 25 mT/m on the imaged object, respectively. The 1D plot of the field and gradient strength maps, along the red, dashed line in the middle of the readout direction, were also shown in (C, D, G, H), respectively.

Figure 2. The reconstructed images and the G maps for the SENSE only accelerated scans (row 1, 3), and the ones with additional local B0 coils modulations (row 2, 4) were compared (both with 200Hz/Pixel). The SENSE reconstruction with 7-fold acceleration starts to contain strong noise amplification, while with additional local coil modulation, only very mild artifacts appeared in the 8-fold reconstruction. As the acceleration factor grew, the encoding efficiency with the local modulation does not drop as quickly as SENSE reconstruction, especially around the peripheral regions.

Figure 3. The reconstructed images and the G maps with the phase offset patterns A and B were compared, given low scanner bandwidth (200Hz/Pixel). As the acceleration factor increases, the SNR efficiency of phase B becomes slightly higher than A. Note that, the SNR loss (i.e., G maps values) in 10kHz given 200Hz/Pixel was generally higher than 7kHz in Figure 2, although the reconstruction accelerated by 6-fold and 7-fold with 10kHz modulations didn't necessarily seem worse.

Figure 4. The reconstructed images with local B0 modulations in varying frequencies, given higher scanner bandwidth (380Hz/Pixel). As the acceleration factor increased, the reconstruction in lower modulation frequency gradually showed advantages in SNR efficiency. The SNR amplification was generally higher compared to identical modulation frequencies given lower scanner bandwidth. Note that such a result is a mixed effect of different encoding patterns and different levels of current damping by the power amplifier.

Figure 5. The G maps for sinusoidal B0 modulations in varying frequencies, given higher scanner bandwidth (380Hz/Pixel). The tendency generally matched with the noise amplification patterns in Figure 4. Most large G map values were concentrated around the central regions of the object, where the modulation strength was the weakest.

Proc. Intl. Soc. Mag. Reson. Med. 31 (2023)
4652
DOI: https://doi.org/10.58530/2023/4652