Visualizing artery-specific blood flow patterns above the circle of Willis with vessel-encoded arterial spin labeling.

Thomas W Okell,James V Byrne,Meritxell Garcia,Michael A Chappell,Peter Jezzard

doi:10.1002/mrm.27507

Abstract

PurposeTo establish the feasibility of using vessel‐encoded pseudocontinuous arterial spin labeling (VEPCASL) for noninvasive vascular territory imaging (VTI) and artery‐specific dynamic angiography of a large number of arterial branches above the circle of Willis within a clinically feasible scan time.Methods3D time‐of‐flight angiography was used to select a labeling plane and establish 7 pairs of encoding cycles. These were used for VEPCASL VTI and dynamic 2D angiography (8 min and 3 min acquisition times, respectively) in healthy volunteers, allowing the separation of signals arising from 13 arterial branches (including extracranial arteries) in postprocessing. To demonstrate the clinical potential of this approach, VEPCASL angiography was also applied in 5 patients with brain arteriovenous malformation (AVM).ResultsIn healthy volunteers, the artery‐specific filling of the vascular tree and resulting perfusion territories were well depicted. SNRs were approximately 5 times higher than those achievable with single‐artery selective methods. Blood supply to the AVMs was well visualized in all cases, showing the main feeding arteries and venous drainage.ConclusionsVEPCASL is a highly efficient method for both VTI and dynamic angiography of a large number of arterial branches, providing a comprehensive picture of vascular flow patterns and the effect on downstream tissue perfusion within an acceptable scan time. Automation of labeling plane and vessel‐encoding selection would improve robustness and efficiency, and further refinement could allow quantitative blood flow measurements to be obtained. This technique shows promise for visualizing the blood supply to lesions and collateral flow patterns.

Highlights

The ability to distinguish the arterial source of blood supply to a specific brain region or lesion is important in a range of diseases, from the study of collateral flow in steno‐occlusive disease[1] to the assessment of feeding arteries in arteriovenous malformations/fistulas (AVMs/AVFs)[2] or brain tumors, such as meningiomas
We aim to demonstrate the feasibility of using Vessel‐encoded pseudocontinuous ASL (VEPCASL) to obtain vessel‐selective dynamic angiograms and vascular territory maps arising from a large number of arterial branches above the circle of Willis using a minimal number of vessel‐encoding cycles in scan times acceptable for compliant patients
Example VEPCASL vascular territory maps and angiographic images from 1 healthy volunteer are shown in Figure 2 and Supporting Information Video S1

Summary

Introduction

The ability to distinguish the arterial source of blood supply to a specific brain region or lesion is important in a range of diseases, from the study of collateral flow in steno‐occlusive disease[1] to the assessment of feeding arteries in arteriovenous malformations/fistulas (AVMs/AVFs)[2] or brain tumors, such as meningiomas. Arterial spin labeling (ASL)[4,5] is an MRI‐based technique that can be used to obtain vessel‐selective information noninvasively and without contrast agents. Single‐artery selective ASL techniques[6,7,8] label blood flowing in individually targeted arterial branches of interest.[9,10,11] Whilst being relatively simple to plan and postprocess, SNR efficiency is reduced compared with conventional nonselective ASL because the labeled blood is restricted to a single artery at a time. Vessel‐encoded pseudocontinuous ASL (VEPCASL) is an alternative approach in which the vessels of interest are encoded by periodically modulating the inversion efficiency across the labeling plane.[12] A series of images acquired with different spatial modulations can be “decoded” to calculate how much blood arises from each feeding artery in each voxel. If an efficient encoding can be performed, SNR efficiency can equal that of a nonselective ASL acquisition.[12,13]

Objectives

Results

Conclusion