Introduction

The higher SNR and longer T₁ at 7T has the potential to increase the SNR of pseudo-continuous ASL (PCASL)¹ perfusion measurements by 2-4x compared to 3T.² This could enable higher spatial resolution imaging³ and robust measurement of white matter perfusion.⁴ However, PCASL is a high-SAR labeling method, so short label durations (LDs) and deadtime to extend the TR are used to mitigate the increase in SAR at 7T.^5–9 This decreases the SNR and number of averages that can be acquired within a given scan duration, reducing the benefit compared to typical 3T protocols.

Previous studies^8,9 optimized the PCASL parameters and background suppression (BGS) pulses to maximize PCASL labeling efficiency (LE) and BGS inversion efficiency (IE) over a large range of off-resonance (±200 Hz), but this came with large increases in SAR which required short LDs and prolonged TRs.

Here, we instead optimize the PCASL parameters and BGS pulses to maximize the SNR efficiency (SNR_eff), such that we maximize the SNR for a given scan time, where $$$SNR_{eff}=\frac{SNR}{\sqrt{TR}}\propto\frac{SNR}{\sqrt{RF_{power}}}$$$, $$$SNR\propto LE_{PCASL}\cdot IE_{BGS}^2$$$, and $$$RF_{power}=\int_0^{TR}|B_1^+(t)|^2\cdot dt\propto \int_0^{TR}SAR\cdot dt$$$. We optimize the PCASL parameters over a much smaller range of off-resonance (±50 Hz) to reduce the need for high B₁⁺ and instead use dynamic B₀-shimming¹⁰ of the labeling plane to reduce off-resonance. Together with VERSE and optimized BGS pulses,¹¹ we achieve superior SNR within a 4-minute scan with shorter TRs.

Methods

The PCASL LE was calculated for a range of PCASL parameters using Bloch simulations: B₁⁺_mean=0.1-2µT, G_max=3-15mT/m, G_mean=0.1-2mT/m, RF spacing=0.5-1.16ms, 50% RF duty-cycle. Simulations assumed pulsatile laminar flow,¹² B₀±50Hz, T₁/T₂=2.1s/0.06s. Minimum-SAR-VERSE was used.¹¹ Parameter combinations that exceeded hardware limitations or perturbed static spins >1.5cm from the labeling plane were excluded. Two sets of PCASL parameters were identified: one which maximized SNR_eff and one which maximized LE (Figure 1).

Two non-selective BGS pulses were considered (Figure 1): a standard hyperbolic-secant (HS-BGS) pulse; and a 25% SAR-reduced VERSE HS pulse with the phase-waveform optimized using optimal-control (OC-BGS).¹¹ In each case, the pulses (HS parameters/phase-waveform) were optimized to maximize IE across B₁⁺±50% and B₀±500Hz using Bloch simulations (T₁/T₂=2.1s/0.06s).

In vivo scans were performed in three volunteers on a Siemens 7T Magnetom Plus with a 8Tx/32Rx head-coil at 1st-level SAR constraints. Saturation-inversion-recovery experiments were used to measure the IE of the BGS pulses. Four-minute white-paper¹³ PCASL data (LD=1.8s, PLD=1.8s, GE-EPI readout, 24 slices, 3.4x3.4x5mm³) were acquired to compare the SNR-efficiencies of the PCASL parameter sets and BGS pulses, where the TR was minimized for each volunteer and scan. Dynamic B₀-shimming of the labeling plane reduced resonance offsets to <50 Hz (results presented elsewhere). Using 3DREAM,¹⁴ the reference voltage was set to accommodate the mean B₁⁺ within the middle EPI slice and the nominal PCASL flip angle was increased to achieve, on average, the desired flip angle within the feeding arteries at the labeling plane.

Results

The optimized PCASL parameters and BGS pulses are described in Figure 1. Although the maxSNR_eff parameters have a lower simulated LE than the maxLE case, the PCASL RF_power was reduced by a factor of 5.5, resulting in a 20% higher simulated SNR_eff. The use of VERSE reduced PCASL RF_power by 23-25%.

The in vivo inversion efficiency of the BGS pulses is shown in Figure 2. The OC-BGS pulse had a 25% lower RF_power and only slightly reduced inversion efficiency, in agreement with simulations (Figure 1). However, the OC-BGS pulse was less robust to very low B₁⁺, as expected from simulations (not shown).

The in vivo SNR_eff results are shown in Figures 3 and 4. To satisfy the scanner's 10s-SAR limit, the PCASL flip angle for the maxLE scan had to be reduced by a factor of 1.61±0.06 (mean±SD). The maxSNR_eff+HS-BGS scan achieved both 41% shorter TRs and 38% higher SNR_eff. With the use of the OC-BGS pulse, the TRs were further shortened, but at the cost of some SNR_eff due to the lower IE compared to the HS-BGS pulse.

Example whole-brain quantified CBF maps are shown in Figure 5.

Discussion

Optimizing for SNR_eff (i.e., balancing LE and SAR), instead of simply aiming to maximizing LE, led to much higher SNR and improved temporal resolution. By using dynamic B₀-shimming, we were able to optimize the PCASL parameters over a smaller range of off-resonances (±50Hz rather than ±200 Hz in^8,9) reducing the need for high B₁⁺_mean and reducing SAR.

We further reduced the TR by using a reduced-SAR OC BGS pulse. However, this came with a reduction in SNR_eff in vivo, in contrast to simulations (not shown), suggesting the IE was lower than expected and requiring further investigation.

Acknowledgements

This study was supported by a Sir Henry Dale Fellowship jointly funded by the Wellcome Trust and the Royal Society (Grant Number 220204/Z/20/Z). The Wellcome Centre for Integrative Neuroimaging is supported by core funding from the Wellcome Trust (203139/Z/16/Z).

References

1. Dai W, Garcia D, De Bazelaire C, Alsop DC. Continuous flow-driven inversion for arterial spin labeling using pulsed radio frequency and gradient fields. Magnetic Resonance in Medicine. 2008;60(6):1488-1497. doi:10.1002/mrm.21790

2. Teeuwisse WM, Webb AG, Van Osch MJP. Arterial spin labeling at ultra-high field: All that glitters is not gold. International Journal of Imaging Systems and Technology. 2010;20(1):62-70. doi:10.1002/ima.20218

3. Shao X, Guo F, Shou Q, et al. Laminar perfusion imaging with zoomed arterial spin labeling at 7 Tesla. NeuroImage. 2021;245:118724. doi:10.1016/j.neuroimage.2021.118724

4. Gardener AG, Jezzard P. Investigating white matter perfusion using optimal sampling strategy arterial spin labeling at 7 Tesla. Magnetic Resonance in Medicine. 2015;73(6):2243-2248. doi:10.1002/mrm.25333

5. Saïb G, Koretsky AP, Talagala SL. Optimization of pseudo‐continuous arterial spin labeling using off‐resonance compensation strategies at 7T. Magn Reson Med. November 2021:mrm.29070. doi:10.1002/mrm.29070

6. Meixner CR, Eisen CK, Schmitter S, et al. Hybrid ‐shimming and gradient adaptions for improved pseudo‐continuous arterial spin labeling at 7 Tesla. Magn Reson Med. August 2021:mrm.28982. doi:10.1002/mrm.28982

7. Boland M, Stirnberg R, Pracht ED, Stöcker T. Robust and SAR-efficient whole-brain pseudo-continuous ASL at 7T. In: Proceedings of the 29th Annual Meeting of the ISMRM, Virtual. ; 2019:4963.

8. Wang K, Ma SJ, Shao X, et al. Optimization of pseudo-continuous arterial spin labeling at 7T with parallel transmission B1 shimming. Magnetic Resonance in Medicine. 2022;87(1):249-262. doi:10.1002/mrm.28988

9. Zhao C, Shao X, Shou Q, et al. Whole-Cerebrum distortion-free three-dimensional pseudo-continuous arterial spin labeling at 7T. NeuroImage. 2023;277:120251. doi:10.1016/j.neuroimage.2023.120251

10. Blamire AM, Rothman DL, Nixon T. Dynamic shim updating: A new approach towards optimized whole brain shimming. Magnetic Resonance in Medicine. 1996;36(1):159-165. doi:10.1002/mrm.1910360125

11. Woods JG, Chiew M, Okell TW. Minimizing SAR for SNR-efficient pseudo-continuous Arterial Spin Labeling at 7T. In: Proceedings of the 31st Annual Meeting of the ISMRM, Toronto, Canada. ; 2023:0371.

12. Zhao L, Vidorreta M, Soman S, Detre JA, Alsop DC. Improving the robustness of pseudo-continuous arterial spin labeling to off-resonance and pulsatile flow velocity. Magnetic Resonance in Medicine. 2017;78(4):1342-1351. doi:10.1002/mrm.26513

13. Alsop DC, Detre JA, Golay X, et al. Recommended implementation of arterial spin-labeled perfusion MRI for clinical applications: A consensus of the ISMRM perfusion study group and the European consortium for ASL in dementia. Magnetic Resonance in Medicine. 2015;73(1):102-116. doi:10.1002/mrm.25197

14. Ehses P, Brenner D, Stirnberg R, Pracht ED, Stöcker T. Whole‐brain B1‐mapping using three‐dimensional DREAM. Magnetic Resonance in Medicine. 2019;(November 2018):mrm.27773. doi:10.1002/mrm.27773

15. Jenkinson M, Beckmann CF, Behrens TEJ, Woolrich MW, Smith SM. FSL. NeuroImage. 2012;62(2):782-790. doi:10.1016/j.neuroimage.2011.09.015