Arterial spin labeling (ASL) is an approach for nonenhanced MR angiography that acquires two image sets which are later subtracted. The image sets differ only in the magnetization states of arterial spins. The process by which this difference in arterial magnetization is imparted is referred to as ‘labeling.’ Subtraction of the image sets displays the labeled arteries and provides complete suppression of background signal.

The first reports of ASL for nonenhanced MR angiography were made in the mid and late 1980s [3–5]. Early implementations of ASL for MR angiography used spin-echo and gradient-echo readouts, but most modern implementations now utilize readouts that are faster and more SNR-efficient. Two popular readouts meeting these criteria are balanced steady-state free precession imaging (bSSFP) and fast spin-echo (FSE) imaging.

As will be explained, labeling of arterial magnetization in ASL MRA can be performed in two ways. Arterial spin labeling can be achieved by either applying RF pulses to alter arterial magnetization, or relying on flow-related dephasing made possible by magnetic field gradients.

Labeling of arterial spins with inversion RF pulses is often considered the most traditional form of arterial spin labeling. This variation of arterial spin labeling is often referred to as ‘pulsed’ ASL. In its most basic form, this method consists of applying an inversion recovery RF pulse, typically of adiabatic type, to arterial spins flowing into the target vasculature, followed by an inversion delay time (TI), and an imaging readout. The difference of this data set and one acquired without the inversion RF pulse selectively displays the labeled arterial spins on a background of no signal. This labeling approach is often referred to as “Signal Targeting with Alternating inversion Recovery” (STAR) [6]. An illustration of STAR ASL MRA is shown in Fig. 21.4a.

Another approach for pulsed arterial spin labeling is “Flow Alternating Inversion Recovery” (FAIR) (Fig. 21.4b) [7]. This approach applies a slice selective inversion RF pulse to the target vasculature in one set and a non-selective inversion RF pulse to the second data set. Subtraction of the data sets displays vascular spins flowing into the target region. Compared with STAR, FAIR ASL may, depending on venous orientation and flow rate, display venous flow into the target region which can be undesirable. As compared to STAR, the FAIR approach better equalizes magnetization transfer effects between the two acquired data sets, which can improve the suppression of background tissue. This magnetization transfer effect, however, is usually not a concern with nonenhanced ASL MRA because of its small magnitude compared to the large signal within arteries.

Unlike pulsed arterial spin labeling, which applies short RF inversion pulses at a fixed TI time before imaging, another class of ASL techniques apply a continuous stream of RF energy for spin labeling. These methods are referred to as “continuous” or “pseudo-continuous”. Whereas continuous arterial spin labeling requires the use of a separate labeling RF coil which limits widespread use in the clinical setting, pseudo-continuous arterial spin labeling (also referred to as “pulsed-continuous” arterial spin labeling) can be applied without additional hardware requirements [8]. An implementation of pseudo-continuous ASL (pCASL) is shown in Fig. 21.4c. Compared to pulsed techniques, pCASL can provide larger arterial signal to noise ratio (SNR) in regions near the pseudo-continuous labeling plane where short effective inversion times are realized. Drawbacks of pCASL include increased RF energy deposition compared with pulsed approaches, and increased sensitivity to off-resonance as well as flow velocity which can reduce labeling efficiency and arterial SNR. Example MR angiograms obtained with pulsed and pseudo-continuous ASL are shown in Fig. 21.5.

Another means for labeling arterial spins during ASL is based on differential flow between systole and diastole in vessels containing pulsatile flow. This approach, originally referred to as gated subtraction angiography [3, 5], subtracts data sets acquired in systole and diastole. Brisk flow during systole results in the dephasing of arterial spins during the echo train and renders arteries dark. Conversely, arterial signal is preserved in diastole due to slow flow. Subtraction displays the arterial anatomy (Fig. 21.6). A commercially available technique for nonenhanced MRA using this labeling scheme is gated subtraction FSE imaging [9], which is also known as “Fresh Blood Imaging” (FBI), “Native Sampling Perfection with Application optimized Contrasts using different flip-angle Evolutions” (Native SPACE), “Delta Flow”, and “Trance”. The flow sensitivity of this technique can be tailored with partial Fourier sampling, adjustment of the refocusing flip angle, additional gradient lobes for flow compensation or spoiling, and by orientation of the frequency encoding axis. Specifically, reduced dephasing and signal loss is achieved with the use of partial Fourier acquisition, larger spin-echo refocusing RF pulse flip angles, flow compensation gradients, and by aligning the phase-encoding axis parallel to the arterial flow direction.

With gated subtraction FSE imaging, dephasing of arterial spins during systole is accomplished by gradients within the imaging readout. Dephasing of arterial spins during systole may also be achieved by applying specially designed magnetization preparations before the imaging readout [10]. Figure 21.7 shows the design of one such sequence. Using RF and gradient pulses, these magnetization preparations spoil arterial signal based on its velocity or acceleration, leaving signal from stationary or non-pulsatile tissue unaltered. These approaches are most frequently applied before imaging readouts, such as bSSFP, which preserve arterial signal even during phases of substantial flow. Names of these techniques include “Flow Sensitive Dephasing” (FSD) and “Acceleration Dependent Vascular Anatomy for Non-Contrast-Enhanced MRA” (ADVANCE-MRA). With these methods, the degree of flow spoiling depends on the timing, duration, and strength of the gradient lobes applied. Gradients may also be applied along multiple axes to enhance flow spoiling and improve angiographic detail.

Main drawbacks of ASL MRA include increased motion sensitivity and doubled scan time compared with non-subtractive nonenhanced MRA techniques. For scans requiring cardiac gating, knowledge of the systolic and diastolic phases of fast and slow flow is important to optimize image quality and avoid artifacts. This information can be obtained with phase contrast imaging or preparatory scans that survey several phases of the cardiac cycle. For gated subtractive techniques, abnormal flow patterns in patients with vascular disease can complicate the acquisition of dark-blood and bright-blood contrasts in systole and diastole, which can lead to image artifacts.

Some ASL methods for MRA require gating while others do not. In particular, ASL techniques based on differential RF pulses often do not require cardiac gating when coupled with readouts that preserve arterial signal during brisk flow, such as bSSFP. This is especially true in vascular beds containing predominantly continuous flow such as the carotid and intracranial arteries. Conversely, techniques that use imaging readouts that are sensitive to fast flow, such as FSE, generally require cardiac gating.

Although not often used in clinical practice, arterial spin labeling can be used to image a variety of vascular beds. In the evaluation of peripheral vascular disease without the use of contrast agents, gated subtractive FSE and flow dephased methods may be used. Of these methods, gated subtractive FSE imaging is the only technique that is commercially available. However, both gated techniques have good reported accuracy for imaging peripheral vascular disease in the distal lower extremities. Both gated techniques have potential for imaging the hands and feet, where contrast-enhanced MRA can be challenging due to slow arterial flow and short arterial to venous transit times.

In the evaluation of intracranial vascular disease, ASL can provide for time-resolved display of arterial flow patterns which may be helpful for assessing the functional significance of various disorders, including arterial occlusions, stenoses, and arteriovenous malformations. This information can be used to supplement the static MR angiograms obtained with 3D TOF MRA. In the carotid arteries, ASL can display arterial stenosis, but more validation is needed to assess the accuracy of the method and the merits and weaknesses of the technique in comparison with 2D and 3D TOF MRA.

Arterial spin-labeled MRA of the renal arteries is feasible but, as with the use of ASL in other vascular territories, large validation studies are lacking. In imaging the renal arteries, there is stiff competition from non-subtractive techniques such as inflow inversion recovery (IFIR) which are faster, less sensitive to motion, and have been better validated. ASL imaging has also been found capable of displaying portal venous anatomy.

Other than TOF and ASL techniques, a few non-subtractive flow-dependent nonenhanced MRA techniques have found increased utilization and development in recent years. Although these techniques still depend on the presence of flow for angiography, as is the case with traditional TOF MR angiography, they differ substantially from TOF in their implementation, timing and display of arterial anatomy. These methods are reviewed here.

Inflow inversion-recovery (IFIR) refers to a technique in which angiographic contrast is obtained by inflow of unsaturated arterial spins into the target vasculature and suppression of background spins through application of an inversion-recovery prepulse [11]. Figure 21.8 shows a timing diagram for such a sequence. Unlike traditional TOF MRA which applies a train of series of largely interrupted RF pulses separated by a short time interval, the IFIR method applies a slab-selective inversion pulse followed by a substantial inversion time (TI). This TI time, which is selected to ensure adequate inflow into the vascular anatomy under interrogation and an adequate degree of background suppression, is followed by a rapid imaging readout with a short interecho spacing.

Stay updated, free articles. Join our Telegram channel

Join