Optimization of pseudo‐continuous arterial spin labeling using off‐resonance compensation strategies at 7T

Abstract

Purpose The sensitivity of pseudo‐continuous arterial spin labeling (PCASL) to off‐resonance effects (ΔB0) is a major limitation at ultra‐high field (≥7T). The aim of this study was to assess the effectiveness of different PCASL ΔB0 compensation methods at 7T and measure the labeling efficiency with off‐resonance correction. Theory and Methods Phase offset errors induced by ΔB0 at the feeding arteries can be compensated by adding an extra radiofrequency (RF) phase increment and transverse gradient blips into the PCASL RF pulse train. The effectiveness of an average field correction (AVGcor), a vessel‐specific field‐map‐based correction (FMcor) and a vessel‐specific prescan‐based correction (PScor) were compared at 7T. After correction, the PCASL labeling efficiency was directly measured in feeding arteries downstream from the labeling location. Results The perfusion signal was more uniform throughout the brain after off‐resonance correction. Whole‐brain average perfusion signal increased by a factor of 2.4, 2.5, and 2.1, respectively, with AVGcor, FMcor and PScor compared to acquisitions without correction. With off‐resonance correction, the maximum labeling efficiency was ~0.68 at mean B1 (B1mean) of 0.70 µT when using a mean gradient (Gmean) of 0.25 mT/m. Conclusion Either a prescan or a field map can be used to correct for off‐resonance effects and retrieve a good brain perfusion signal at 7T. Although the three methods performed well in this study, FMcor may be better suited for patient studies because it accounted for vessel‐specific ΔB0 variations. Further improvements in image quality will be possible by optimizing the labeling efficiency with advanced hardware and software while satisfying specific absorption rate constraints.

Whole-Cerebrum distortion-free three-dimensional pseudo-continuous arterial spin labeling at 7T

“…Future study comparing 7T ASL with 3T ASL and PET are needed to derive these parameters from well-established reference values. The CBF values listed in Table 1 are comparable to reported values in some previous studies ( Binnie et al, 2022 ; Saïb et al, 2022 ; Tong et al, 2020 ; Wang et al, 2013 ), but are lower than those of a few other studies ( Almeida et al, 2018 ; Parkes et al, 2004 ).…”

“…In addition, a lower G ave of 0.4 mT/m achieved a higher mean IE both in simulations and in vivo than the original G ave of 0.6 mT/m. To mitigate the influence of B0 offset, compensating gradients during pCASL was also proposed without shortening RF spacing and RF duration ( Saïb et al, 2022 ). However, this method requires additional pre-scans and patient-specific adjustments.…”

Whole-Cerebrum distortion-free three-dimensional pseudo-Continuous Arterial Spin Labeling at 7T

“…In addition, a lower Gave of 0.4 mT/m achieved a higher mean IE both in simulations and in vivo than the original Gave of 0.6 mT/m. To mitigate the influence of B0 offset, compensating gradients during pCASL was also proposed without shortening RF spacing and RF duration (Saïb et al, 2022). However, this method requires additional pre-scans and patient-specific adjustments.…”