This educational talk covers the data acquisition of arterial spin labeling (ASL) technique and steps to obtain quantified cerebral blood flow (CBF) map. ASL data acquisition is usually composed of three components: labeling module to tag blood water in major feeding arteries, post-labeling delay module to allow tagged blood to reach tissue, and readout module to acquire images. ASL data processing involves motion correction and perfusion quantification using general kinetic model. In addition, common issues met in practice and examples of clinical application of ASL are briefly discussed.
Labeling module: the two widely used labeling schemes are: pulsed ASL (PASL) and pseudo-continuous ASL (pCASL) [3, 4]. PASL uses a very short pulse (~10 ms) to invert a thick slab (~100 mm) of arterial water spins. It has the advantage of high labeling efficiency and low SAR. While pCASL applies a series of small-flip-angle RF pulses and gradients (this is why this scheme is referred to as pseudo-continuous) that lasts typically ~2 seconds to invert the blood water spins passing through the thin labeling plane. This labeling scheme provides higher SNR than PASL yet higher SAR. Due to the fact that RF transmitter in clinical MRI could not offer continuous RF field with long duration, true continuous ASL is usually not feasible.
Post-labeling delay module: The waiting time (~2 seconds) after labeling, post-labeling delay (PLD), allows the labeled water spins to travel from major feeding arteries to capillary bed [5]. The selection of proper PLD value is a tread-off between minimizing arterial artifacts and retain signal from labeled water spins that decays at the rate of 1/T1. If PLD is too short, most of the labeled spins still reside in large arteries; while if PLD to too long, the final CBF image will be of less signal because the inverted labeled spins has recovered to similar level with control spins.
Readout module: Perfusion signal is usually at the level of ~0.5% of fully recovered magnetization. The signal intensity is also decaying while readout. Therefore, readout sequence with high acquisition efficiency and high SNR is preferred in ASL. 3D segmented acquisition, such as Gradient Spin Echo (GRASE) and stack of spirals, are commonly used [6, 7]. On the other hand, 2D multi-slice acquisition, such as EPI, is also widely used for the brain due to their availability on most MRI scanners [8].
Other advanced techniques: Besides the widely used PASL and pCASL, there are several variations of ASL that serves different imaging purpose. Multi-delay ASL acquires images at several delay time points to obtain temporal information of bolus arrival time [9]. Time-encoded ASL utilizes the flexible block design of labeling module to achieve similar results yet with increase acquisition efficiency [10]. Vessel-selective ASL allows flow territory mapping [11]. Velocity-selective ASL tags flowing water spins and thus enables larger amount of labeled bolus [12]. Most recently, when combining with MR fingerprinting, several parametric maps of hemodynamics can be obtained within one ASL scan [13, 14].
Pre-processing: Due to the intrinsic low SNR of ASL signal, averaging across repeated acquisitions of control and label images is often necessary. Motion between images need to be corrected to align control and label volumes, such that their subtraction and average is free of bright or dark rings at brain boundary.
CBF quantification: The difference image of control and label is proportional to CBF. In order to obtain quantified values, difference image is divided by a proton density image acquired in a separate scan to normalize the signal intensity. The normalized difference image along with assumed constant parameters are then entered into a formula to calculate CBF map, where the formula is derived from general kinetic model [2].
Group statistical analysis: After obtaining CBF maps, region-of-interest (ROI) values can be extracted when overlaying segmentation result of anatomical image (e.g. T1 weighted) to the perfusion maps. These outcome variables can then be readily used for study-specific statistical analyses.
Available toolboxes: There are several existing software tools that implemented the ASL processing steps above, such as ASLtbx, BASIL, ASAP [15-17]. Notably, a cloud-computing based tool, ASL-MRICloud, has been deployed online recently and is capable of automated processing [18]. User only needs a web browser to upload data and download results.
1. Williams, D.S., et al., Magnetic-Resonance-Imaging of Perfusion Using Spin Inversion of Arterial Water. Proceedings of the National Academy of Sciences of the United States of America, 1992. 89(1): p. 212-216.
2. Alsop, D.C., et al., Recommended Implementation of Arterial Spin-Labeled Perfusion MRI for Clinical Applications: A Consensus of the ISMRM Perfusion Study Group and the European Consortium for ASL in Dementia. Magnetic Resonance in Medicine, 2015. 73(1): p. 102-116.
3. Edelman, R.R., et al., Qualitative Mapping of Cerebral Blood-Flow and Functional Localization with Echo-Planar Mr-Imaging and Signal Targeting with Alternating Radio-Frequency. Radiology, 1994. 192(2): p. 513-520.
4. Dai, W.Y., et al., Continuous Flow-Driven Inversion for Arterial Spin Labeling Using Pulsed Radio Frequency and Gradient Fields. Magnetic Resonance in Medicine, 2008. 60(6): p. 1488-1497.
5. Alsop, D.C. and J.A. Detre, Reduced transit-time sensitivity in noninvasive magnetic resonance imaging of human cerebral blood flow. J Cereb Blood Flow Metab, 1996. 16(6): p. 1236-49.
6. Gunther, M., K. Oshio, and D.A. Feinberg, Single-shot 3D imaging techniques improve arterial spin labeling perfusion measurements. Magn Reson Med, 2005. 54(2): p. 491-8.
7. Ye, F.Q., et al., Noise reduction in 3D perfusion imaging by attenuating the static signal in arterial spin tagging (ASSIST). Magn Reson Med, 2000. 44(1): p. 92-100.
8. Jahng, G.H., M.W. Weiner, and N. Schuff, Improved arterial spin labeling method: applications for measurements of cerebral blood flow in human brain at high magnetic field MRI. Med Phys, 2007. 34(11): p. 4519-25.
9. Gunther, M., M. Bock, and L.R. Schad, Arterial spin labeling in combination with a look-locker sampling strategy: inflow turbo-sampling EPI-FAIR (ITS-FAIR). Magn Reson Med, 2001. 46(5): p. 974-84.
10. Wells, J.A., et al., In vivo Hadamard encoded continuous arterial spin labeling (H-CASL). Magn Reson Med, 2010. 63(4): p. 1111-8.
11. Helle, M., et al., Superselective pseudocontinuous arterial spin labeling. Magn Reson Med, 2010. 64(3): p. 777-86.
12. Wong, E.C., et al., Velocity-selective arterial spin labeling. Magn Reson Med, 2006. 55(6): p. 1334-41.
13. Su, P., et al., Multiparametric estimation of brain hemodynamics with MR fingerprinting ASL. Magn Reson Med, 2017. 78(5): p. 1812-1823.
14. Wright, K.L., et al., Estimation of perfusion properties with MR Fingerprinting Arterial Spin Labeling. Magn Reson Imaging, 2018. 50: p. 68-77.
15. Wang, Z., et al., Empirical optimization of ASL data analysis using an ASL data processing toolbox: ASLtbx. Magnetic Resonance Imaging, 2008. 26(2): p. 261-269.
16. Chappell, M.A., et al., Variational Bayesian Inference for a Nonlinear Forward Model. Ieee Transactions on Signal Processing, 2009. 57(1): p. 223-236.
17. Abad, V.M., et al., ASAP (Automatic Software for ASL Processing): A toolbox for processing Arterial Spin Labeling images. Magnetic Resonance Imaging, 2016. 34(3): p. 334-344.
18. Li, Y., et al., ASL-MRICloud: An online tool for the processing of ASL MRI data. NMR Biomed, 2019. 32(2): p. e4051.
19. Li, Y., et al., Cardiac-triggered pseudo-continuous arterial-spin-labeling: A cost-effective scheme to further enhance the reliability of arterial-spin-labeling MRI. Magn Reson Med, 2018. 80(3): p. 969-975.
20. Grade, M., et al., A neuroradiologist's guide to arterial spin labeling MRI in clinical practice. Neuroradiology, 2015. 57(12): p. 1181-1202.