1273

UTE-Based DW-SSFP MRI for 7T

Kwan-Jin Jung¹
¹Beckman Institute, Biomedical Imaging Center, University of Illinois at Urbana-Champaign, Urbana, IL, United States

Synopsis

Keywords: Diffusion Acquisition, Diffusion Tensor Imaging, DW-SSFP

Motivation: At 7T the conventional spin-echo EPI diffusion sequence suffers from the B1+ inhomogeneity and geometric distortion due to refocusing RF pulses and EPI readout.

Goal(s): To develop a diffusion imaging sequence without refocusing RF pulses and EPI readout at 7T.

Approach: Develop a 3-dimensional DW-SSFP sequence with a spiral readout to reduce the geometric distortion, to maximize the diffusion gradient time, and to reduce susceptibility effect.

Results: The proposed DW-SSFP sequence was successful in reducing the B1+ inhomogeneity, the geometric distortion, and the susceptibility effect compared to the spin-echo EPI diffusion sequence using cadaveric head specimens at 7T.

Impact: This sequence enables the acquisition of high-resolution diffusion images that do not suffer from the B1+ inhomogeneity and geometric distortion often observed at 7T. It provides good quality fiber tracts and fractional anisotropy maps of the brain.

Introduction

A spin-echo-based diffusion imaging sequence suffers from RF nonuniformity due to refocusing 180° RF pulses at 7T ¹. In contrast, a diffusion-weighted steady state free presession (DW-SSFP) sequence uses a small flip angle RF pulse ^2,3 that reduces the RF nonuniformity in images at 7T. On the other hand, the EPI readout in diffusion imaging results in not only geometric distortions but also an increased echo shift from the echo center in DW-SSFP ⁴. These effects increase at 7T compared to 3T. Therefore, DW-SSFP sequence was used to reduce the RF nonuniformity and the spiral-based ultra-short-echo-time (UTE) readout was used to reduce the geometric distortion and the echo shift time ⁵. We demonstrated the UTE-based DW-SSFP sequence on four cadaveric head specimens at 7T.

Methods

The DW-SSFP sequence with the 3D spiral trajectory is shown in Figure 1. The echo shift (ΔTE) is minimized compared to the EPI readout ⁶, which allows a longer duration of the diffusion encoding gradient (G_DW). One volume of b0 was obtained with b=3.2 s/mm² to spoil the FID component in SSFP ⁷, followed by 6 diffusion-weighted acquisitions for 6 directions of MDDW with b=1420 s/mm². The b values were for the echo signal from the immediate prior TR. The scan parameters were: TR=25ms, echo shift (ΔTE)=0.17ms, voxel=2mm isotropic, duration of G_DW=20ms, amplitude of G_DW=37mT/m, flip angle=30°, acceleration=2, and scan time=5:56. As a comparison, a multi-slice spin-echo EPI diffusion sequence with a bipolar diffusion gradient was used for DW-EPI with b=2000 s/mm² in 24 directions and 4 interleaved b0: TR/TE=6300/88ms, voxel=2mm isotropic, and scan time=3:47. Four cadaveric head specimens were acquired in a fresh and never-frozen state from an external lab (Science Care, Coral Spring, FL) and they were kept at 4°C before MRI scans. The specimens had been flushed with cephalus-formalin. The four specimens were scanned at 7T: two with a single channel transmission (STX) head RF coil and the other two with an 8-ch parallel transmission (PTX) head RF coil. Additionally, anatomical images of 3-dimensional T1W (MP2RAGE) and T2W (SPACE) were collected. The MP2RAGE images were denoised using a lossless algorithm and normalized for the B1 sensitivity ⁸. The universal RF pulses were used for the T2W SPACE scan with the PTX coil ⁹. The fractional anisotropy (FA) was estimated using DTIFIT in FSL ¹⁰. The reduced geometric distortion of the DW-SSFP sequence was demonstrated by tracking the corticospinal tracts using DSI Studio ¹¹.

Results

Results from one representative specimen (age=82, gender=F) is shown in Figures 2, 3, and 4. All four specimens (average age=72, 1 female) are shown in Figure 5 for T2W images. The reduced geometric distortion of the DW-SSFP sequence was demonstrated in the sagittal view of b0 images (Figure 2). The signal reduction from the susceptibility effect was minimum in the DW-SSFP sequence due to the short echo shift. In this figure, the RF nonuniformity from 180° RF pulses of the DW-EPI sequence reduced the signal at the head center region. The reduced signal at the head center resulted in noisy and corrupted FA estimates in Figure 3. In contrast, the signal at the head center region was uniform and hence FA was not corrupted in the DW-SSFP sequence. Furthermore, the reduced geometric distortion in the DW-SSFP sequence enabled a good tracking of the corticospinal tracts which were not well detected in the DW-EPI sequence (Figure 4). The signal void at the head center was also observed in the T2W-SPACE images which were obtained from refocused echoes (Figure 5). Interestingly, the signal nonuniformity was worse with the Universal RF pulse of the PTX than the STX RF pulse.

Discussion

The reduction of signals at the center of the specimens in DW-EPI and T2W-SPACE sequences was less pronounced in healthy volunteers. This may suggest that the B1+ field was altered by the specimen’s lower temperature. Diffusion-weighting b value and diffusivity depend on relaxation times and flip angle in the DW-SSFP sequence ^{12, 13}. While the DW-SSFP sequence provides valid and enhanced FA and tracts ⁶, it is necessary to address the motion effects seen in in vivo scans ¹⁴. The motion sensitivity can be reduced by using a flow-compensated diffusion gradient and motion navigator ¹⁵.

Conclusion

The proposed DW-SSFP sequence with UTE was effective in reducing the RF nonuniformity and geometric distortion in diffusion imaging at 7T. This advantage of the DW-SSFP sequence contributed to significantly improved estimates of FA over the whole brain and successful tracking of corticospinal tracts.

Acknowledgements

The scanning portion of this work was supported by the Beckman Institute’s Biomedical Imaging Center at the University of Illinois at Urbana-Champaign. The cadaveric specimens were provided by Dr. Tracey Wszalek prior to their use in a neurosurgical dissection course for neurosurgical residents. The author appreciates the specimen preparation by Debbie Styan and the scan assistance by medical students from the Carle-Illinois College of Medicine.

References

1. Wu X, Auerbach EJ, Vu AT, Moeller S, Lenglet C, Schmitter S et al. High-resolution whole-brain diffusion MRI at 7T using radiofrequency parallel transmission. Magn Reson Med. 2018;80(5):1857-70.

2. Le Bihan D. Intravoxel incoherent motion imaging using steady-state free precession. Magn Reson Med. 1988;7(3):346-51.

3. McNab JA, Jbabdi S, Deoni SC, Douaud G, Behrens TE, Miller KL. High resolution diffusion-weighted imaging in fixed human brain using diffusion-weighted steady state free precession. Neuroimage. 2009;46(3):775-85.

4. McNab JA, Miller KL. Sensitivity of diffusion weighted steady state free precession to anisotropic diffusion. Magn Reson Med. 2008;60(2):405-13.

5. Jung KJ, Sutton B, editors. Three-Dimensional Sodium MRI Using A Rotation of Spiral Disc (RSD) Trajectory. Int Soc Magn Reson Med; 2021.

6. Miller KL, McNab JA, Jbabdi S, Douaud G. Diffusion tractography of post-mortem human brains: optimization and comparison of spin echo and steady-state free precession techniques. Neuroimage. 2012;59(3):2284-97.

7. Jung KJ. Synthesis methods of multiple phase-cycled SSFP images to reduce the band artifact and noise more reliably. Magn Reson Imaging. 2010;28(1):103-18.

8. Jung K-J, editor. Denoising of 7T MP2RAGE MRI with Preserving the Brain Signal. Int Soc Magn Reson Med; 2023; Toronto, CA.

9. Gras V, Boland M, Vignaud A, Ferrand G, Amadon A, Mauconduit F et al. Homogeneous non-selective and slice-selective parallel-transmit excitations at 7 Tesla with universal pulses: A validation study on two commercial RF coils. Plos One. 2017;12(8).

10. Behrens TEJ, Woolrich MW, Jenkinson M, Johansen-Berg H, Nunes RG, Clare S et al. Characterization and propagation of uncertainty in diffusion-weighted MR imaging. Magnet Reson Med. 2003;50(5):1077-88.

11. Yeh FC. Shape analysis of the human association pathways. Neuroimage. 2020;223:117329.

12. Tendler BC, Foxley S, Cottaar M, Jbabdi S, Miller KL. Modeling an equivalent b-value in diffusion-weighted steady-state free precession. Magn Reson Med. 2020;84(2):873-84.

13. McNab JA, Miller KL. Steady-state diffusion-weighted imaging: theory, acquisition and analysis. NMR Biomed. 2010;23(7):781-93.

14. Jung KJ, Cho ZH. Reduction of Flow Artifacts in Nmr Diffusion Imaging Using View-Angle Tilted Line-Integral Projection Reconstruction. Magnet Reson Med. 1991;19(2):349-60.

15. McNab JA, Gallichan D, Miller KL. 3D Steady-State Diffusion-Weighted Imaging With Trajectory Using Radially Batched Internal Navigator Echoes (TURBINE). Magnet Reson Med. 2010;63(1):235-42.

Figures

Figure 1. DW-SSFP sequence diagrams. (A) One TR period of the sequence diagram is shown. It is a 3D sequence with a rectangular RF pulse. The readout is a spiral-in trajectory to minimize the echo shift (ΔTE). The rewind gradient (G_Rew) is to refocus the spiral readout in each TR. The diffusion encoding gradient (G_DW) is applied on x and z axes in this example figure and they are rotated for different diffusion encoding directions. (B) The spiral leaves are interleaved to form a 2D K-sapce disc and the disc was rotated to fill the 3D K-sapce.

Figure 2. A comparison of b0 images between the DW-EPI and DW-SSFP sequences in reference to the T1W image in three orthogonal planes. The geometric distortion is clearly shown in the sagittal view (the first columns). The RF nonuniformity effect in the DW-EPI sequence resulted in signal void at the head center region.

Figure 3. A comparsion of FA maps from the DW-EPI and DW-SSFP sequences. The psuedo-colored FA maps were overlaid on the T1W images in three orthogonal planes. The head center region was corrupted due to the signal void in the DW-EPI sequence, which was not seen in the DW-SSFP sequence.

Figure 4. A comparison of corticospine tracts from the DW-EPI and DW-SSFP sequences. The severe geometric distortion at the brain stem terminated the tracts in the DW-EPI sequence. In contrast, the corticospine tracts were successfully tracked in the DW-SSFP sequence.

Figure 5. T2W SPACE images acquired from four cadaveric cephalus specimens. The two specimens in the upper row (A and B) were scanned with a STX coil, while the other two specimens in the bottom row (C and D) were scanned with the PTX coil using the Universal RF pulse. The signal void due to the RF nonuniformity were more pronounced at the PTX images, which suggested that the universal RF pulse might not be effective when the cadaveric specimens were 4° C in temperature.

Proc. Intl. Soc. Mag. Reson. Med. 32 (2024)

1273

DOI: https://doi.org/10.58530/2024/1273