Three-dimensional acquisition of cerebral blood volume and flow responses during functional stimulation in a single scan

Abstract

In addition to the BOLD scan, quantitative functional MRI studies require measurement of both cerebral blood volume (CBV) and flow (CBF) dynamics. The ability to detect CBV and CBF responses in a single additional scan would shorten the total scan time and reduce temporal variations. Several approaches for simultaneous CBV and CBF measurement during functional MRI experiments have been proposed in two-dimensional (2D) mode covering one to three slices in one repetition time (TR). Here, we extended the principles from previous work and present a three-dimensional (3D) whole-brain MRI approach that combines the vascular-space-occupancy (VASO) and flow-sensitive alternating inversion recovery (FAIR) arterial spin labeling (ASL) techniques, allowing the measurement of CBV and CBF dynamics, respectively, in a single scan. 3D acquisitions are complicated for such a scan combination as the time to null blood signal during a steady state needs to be known. We estimated this using Bloch simulations and demonstrate that the resulting 3D acquisition can detect activation patterns and relative signal changes of quality comparable to that of the original separate scans. The same was found for temporal signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). This approach provides improved acquisition efficiency when both CBV and CBF responses need to be monitored during a functional task.

Keywords: ASL; Arterial spin labeling; CBF; CBV; FAIR; Flow-sensitive alternating inversion recovery; Three-dimensional; VASO; Vascular-space-occupancy; Whole-brain; fMRI.

Figure 1

Pulse sequence of the combined 3D VASO-FAIR approach. A pair of interleaving slice-selective (SS) and nonselective (NS) scans are shown. A magnetization transfer (MT) prepulse is added before the adiabatic FOCI inversion pulses. The imaging module used here is a 3D fast GRE readout for both VASO and FAIR ASL images, in which VASO signal is acquired at blood nulling time TI₁ and ASL signal at time TI₂. A post-saturation module comprising of a non-selective 90° saturation pulse and spoiler gradients is applied immediately after the FAIR ASL readout.

Figure 2

Bloch simulations of the signal evolution for blood in (blue) and outside (red) the inversion slab applied in the slice-selective (SS) scan with (a) and without (b) the post-saturation module. The solid vertical black lines indicate the inversion pulses (labeled as “inv”) in the interleaving slice-selective (SS) and non-selective (NS) scans. The short dashed vertical lines represent the first excitation pulses (labeled as “ex”) in each image acquisition module, the durations of which are labeled in light shade. The MT period before the inversion pulse is labeled in dark shade. Steady state blood nulling time for VASO is marked as TI₁ (solid vertical green lines) and the post-labeling delay for FAIR as TI₂. (a) With the post-saturation module (labeled as “post-sat”) applied immediately after the second readout, a steady state is built after the first TR, at which both blood in and outside the SS inversion slab is nulled at TI₁ in the NS scan. (b) When the post-saturation module is not applied, it takes three TRs for both in-slab and out-of-slab blood to reach steady state (for the experimental parameters used) and the out-of-slab blood is not properly nulled even at steady state (zoomed inlet).

Figure 3

Representative images from one subject. (a) Typical VASO images (3D fast GRE readout, 16 slices) from the 3D VASO-FAIR sequence. (b) Typical FAIR ASL difference maps (perfusion weighted images) in 3D VASO-FAIR. (c) Corresponding VASO images in the separate VASO scan. (d) Corresponding perfusion weighted images in the separate FAIR ASL scan. Images in (a) and (c), and (b) and (d) are on the same scales, respectively.

Figure 4

Comparison of functional MRI results between the combined and separate scans. Representative activation maps from one subject for (a) separate VASO scan, (b) VASO in the combined 3D VASO-FAIR method; (d) separate FAIR ASL scan, and (e) FAIR ASL in the 3D VASO-FAIR scan are shown. The activated voxels are highlighted with their t-scores. Time courses averaged over common voxels activated in both separate (black circle) and combined (red triangle) scans (averaged over subjects, n = 6) for VASO and FAIR are shown in (c) and (f), respectively. Error bars represent inter-subject variations. The vertical dotted lines indicate the start and end of visual stimulation. Four blocks of baseline and stimulation periods were averaged to one block.