Elsevier

Magnetic Resonance Imaging

Volume 73, November 2020, Pages 15-22
Magnetic Resonance Imaging

Technical Note
Improved selective visualization of internal and external carotid artery in 4D-MR angiography based on super-selective pseudo-continuous arterial spin labeling combined with CENTRA-keyhole and view-sharing (4D-S-PACK)

https://doi.org/10.1016/j.mri.2020.07.013Get rights and content

Abstract

Purpose

Four-dimensional magnetic resonance angiography (4D-MRA) based on super-selective pseudo-continuous arterial spin labeling, combined with Keyhole and View-sharing (4D-S-PACK) was introduced for scan-accelerated vessel-selective 4D-MRA. Label selectivity and visualization effectiveness were assessed.

Methods

Nine healthy volunteers were included in the study. The label selectivity for the imaging of internal carotid artery (ICA) and external carotid artery (ECA) circulation was assessed qualitatively. The contrast-to-noise ratio (CNR) in 4D-S-PACK was measured in four middle cerebral artery (MCA) and superficial temporal artery (STA) segments and compared with that in contrast-inherent inflow-enhanced multi-phase angiography combined with the vessel-selective arterial spin labeling technique (CINEMA-select). Vessel-selective arterial visualization in 4D-S-PACK was assessed qualitatively in a patient with dural arteriovenous fistula and compared with digital subtraction angiography (DSA) and non-vessel selective 4D-PACK.

Results

4D-S-PACK vessel selectivity was judged to be at a clinically acceptable level in all cases except one ECA-targeted label. The CNR was significantly higher using 4D-S-PACK compared with CINEMA-select in MCA and STA peripheral segments (p < 0.001). In patient examination, territorial flow visualization in feeding artery and draining vein circulation on 4D-S-PACK were comparable with that on DSA and the identification of such responsible vessels was easier on 4D-S-PACK than on 4D-PACK.

Conclusion

4D-S-PACK showed high vessel-selectivity and higher visualization effectiveness compared with CINEMA-select. One clinical case was performed and ICA and ECA territorial flow was successfully visualized separately, suggesting clinical usefulness.

Introduction

Digital subtraction angiography (DSA) is the current standard diagnostic modality for neurovascular disease. It provides anatomical and flow dynamic information of vessels with high spatial and temporal resolution. However, this method carries risks of neurological complications, adverse reactions associated with contrast agents and radiation exposure [1].

As a non-invasive approach, arterial spin labeling (ASL) demonstrated the ability of non-contrast enhanced intracranial magnetic resonance angiography (MRA) with three dimensional (3D) spatial and dynamic acquisition (4D) in such neurovascular diseases as steno-occlusive disease, moyamoya disease, arteriovenous malformations (AVM) and dural arteriovenous fistula (dAVF) [[2], [3], [4], [5], [6], [7]].

Furthermore, there are reports of 4D-MRA techniques aimed at vessel-selective flow visualization, which is another clinically important feature of DSA [[8], [9], [10], [11]]. Understanding territorial flow of each vessel is important in determining the treatment strategy, such as identification of feeding arteries, detailed preoperative evaluations, and planning of interventional or surgical procedures. One vessel-selective 4D-MRA approach that has already been clinically validated is contrast inherent inflow-enhanced multi-phase angiography combined with vessel-selective arterial spin labeling technique (CINEMA-select) [9]. This approach has been shown to have clinical usefulness in the evaluation of intracranial AVMs by detecting the feeding arteries and in the estimation of detailed information about the nidus.

However, as a Look-Locker sampling strategy is used in CINEMA-select to acquire multiple time point data, signal saturation due to consecutive radiofrequency (RF) pulses can lead to insufficient vessel visualization, particularly at later time points. Moreover, CINEMA-select is limited in terms of vessel selectivity. For example, it is difficult for CINEMA-select to separately visualize territorial flow of the internal carotid artery (ICA) and the external carotid artery (ECA), as a spatially non-selective label pulse is used in the foot-to-head direction and the label slab typically includes the common carotid artery [9]. This could be clinically significant in understanding the vessel responsible for extracranial-intracranial bypass flow, the feeding artery, the nidus and the draining vein circulation in AVM or the shunt flow in dAVF.

Four dimensional (4D) MR angiography based on pseudo-continuous arterial spin labeling (pCASL) combined with CENTRA-keyhole and view-sharing, named 4D-PACK, has been proposed as a 4D-MRA technique [2,12]. With this technique, acquisition times were accelerated by using keyhole and view-sharing and better peripheral artery visualization with longer transit times compared to CINEMA was achieved [2,12]. However, this technique does not have vessel selectivity.

Recently, the super-selective technique, a pCASL-based technique, was introduced as a vessel-selective labeling technique [13]. In this study, we propose combining 4D-PACK with super-selective labelling, referred to as 4D-S-PACK (4D-MRA based on super-selective pCASL with CENTRA-keyhole) hereafter. The aim of this study is to investigate 4D-S-PACK label selectivity for ICA and ECA, as well as its capability to visualize individual territorial flow, especially focusing on peripheral flow in patients with delayed or long-transit time flow circulation.

Section snippets

4D-S-PACK

The 4D-S-PACK scheme is visualized in Fig. 1. The concept and scheme of 4D-S-PACK is the same as of 4D-PACK except that the super-selective pulse is used in the pCASL module in the case of 4D-S-PACK [12,13]. In addition to the non-selective pCASL that uses gradient Gz in the direction of the blood flow, gradients perpendicular to the flow direction (Gx, Gy) are added in 4D-S-PACK (Fig. 1a). By dynamically changing the direction of the additional gradients, a circular or elliptical label spot

Results

Representative 4D-MRA images in the right ICA circulation generated from 4D-S-PACK and CINEMA-select are shown in coronal and axial orientations in Fig. 3. The images obtained by 4D-S-PACK exhibited high-flow visualization in peripheral areas (yellow arrows) for MCA. The ICA circulation was visualized without visualization of ECA territorial flow in 4D-S-PACK; ECA territorial flow was visualized in CINEMA-select (blue arrows), as it utilizes common carotid artery labels.

Representative right ECA

Discussion

In this study, we propose a new vessel-selective 4D-MRA approach called 4D-S-PACK, a 4D-MRA technique with super-selective combined pCASL, accelerated using keyhole and view-sharing techniques.

In the examination of the healthy volunteers, 4D-S-PACK showed better artery visualization in the MCA and STA peripheral regions compared to CINEMA-select. In a previous study comparing 4D-PACK and CINEMA, similar results were reported [12]. Though there is a difference between vessel-selective and

Acknowledgements

We thank Tetsuo Ogino, Yuta Akamine and Yu Ueda for valuable help with the imaging sequence development or for giving advice about study design.

References (16)

There are more references available in the full text version of this article.

Cited by (5)

  • Optimization of 4D-MR angiography based on superselective pseudo-continuous arterial spin labeling combined with CENTRA-keyhole and view-sharing (4D-S-PACK) for vessel-selective visualization of the internal carotid artery and vertebrobasilar artery systems

    2022, Magnetic Resonance Imaging
    Citation Excerpt :

    This study was approved by the Institutional Review Board of Kyushu University Hospital (No. 20192029), and written informed consent was obtained from all subjects. [R2.1]: This study used a new approach for vessel-selective 4D-MRA as identical superselective sequence and 4D-S-PACK scheme reported by Obara et al. [18]. In 4D-S-PACK, superselective 4D-MRA images are acquired by dynamically changing the labeling duration and, after the application of a control or label pulse, implementing a shot post-labeling delay (PLD) and then acquiring images using RF and gradient-spoiled T1-enhanced turbo field echo (T1-TFE) acquisition.

View full text