Accelerated territorial arterial spin labeling based on shared rotating control acquisition: an observer study for validation

Abstract

Introduction

Shared rotating control acquisition can shorten the imaging time of territorial arterial spin labeling (tASL) by 33% compared with the normal control acquisition scheme but potentially results in an inaccurate estimate of vascular territories due to imperfect magnetization transfer compensation. Our purpose was to validate the accuracy of the shared rotating control acquisition method in evaluation of vascular territories.

Methods

Twenty-four patients underwent tASL at a 3.0-T MRI with the conventional normal control acquisition method. Composite vascular territory maps, in which the blood flows from the right and left internal carotid arteries and the posterior circulation were encoded in red–green–blue, were generated as a normal averaged control-label scheme and as a simulated shared rotating control scheme. Two observers independently reported the most dominant territorial flow in 26 brain regions corresponding to the arterial segments at three post-labeling time points. Inter-reader and inter-method agreements were analyzed using κ statistics.

Results

Overall inter-reader agreements were excellent for both the normal control and the shared rotating control methods (κ = 0.98, respectively). Overall inter-method agreement was also excellent (κ = 0.98), although relatively low agreement was noted in the bilateral posterior cerebral artery territories (κ = 0.79 to 0.93).

Conclusion

Our results suggested that tASL using shared rotating control acquisition can provide information on the vascular territories comparable to that obtained using the normal control acquisition while substantially shortening the imaging time.