Arterial spin labeling (ASL) permits the estimation of cerebral blood flow (CBF) with magnetic resonance imaging (MRI) non-invasively, using magnetically "labelled" (or "tagged") arterial blood water spins as endogenous traces, instead of exogenously administered tracer (i. e. contrast media). To accomplish this, the longitudinal magnetization of arterial blood water is manipulated so that it differs from the tissue magnetization, thereby the magnetic labelling effect exponentially decays with T1 time constant. In practice, CBF is determined from whole-brain echo-planar imaging (EPI) time-series by taking the signal intensity differences between consecutaive volumes acquired without and with tagging (dM), thereby subtracting out the static magnetization of the imaging slab. Besides obtaining a signal proportial to CBF (relative CBF), this (subtle) difference can be futher modeled to derive a quantitative (absolute) CBF image showing perfusion in ml/100g/min at each voxel.
In a typical ASL image time-series, subtraction of the tag from the preceeding control image results in an image with intensity proportional to CBF. The most basic way to perform control-tag calculations would be to directly subtract tag from control images. However, since the control and tag images are not acquired at exactly the same time point, a "surround" subtraction, i. e. computing the difference between each image and the average of its two nearest neighbors, would minimize these signal artifacts due to this difference in timing. Moreover, this would further reduce the contamination of the BOLD effect on the resulting perfusion-weighted time-series.
An even better approach, gathering the same advantages of surrounding subtraction but without sacrificing the temporal resolution of the resulting time-seres, consists in (i) separating the control and tag time-series, (ii) temporally interpolating both time-series at the original TR, and (iii) directly subtracting the two interpolated time-series at exactly corresponding time points. In addition, by taking, besides the difference, the sum of the (interpolated) control and tag time-series, a standard T2*-weighted time-series, i. e. with "pure" BOLD weighting effect, can be reconstructed, allowing for comparative perfusion (dM-weighted) and BOLD (T2*-weighted) dynamic imaging and analysis. This is the approach used in the ASL plugin.
ASL perfusion can be accomplished using a variety of approaches (and not only for the brain), but the most common ASL approaches use either pulsed labeling with an instantaneous spatially selective saturation or inversion pulse or continuous labeling, most typically by flow driven adiabatic fast passage. Whereas continuous ASL (CASL) sequences use continuous radio-frequency (RF) irradiation for blood water tagging, pulsed ASL (PASL) sequences use short adiabatic pulses (i.e. 10 ms long) to tag blood spins. PASL is associated with lower contrast-to-noise ratios (CNR) than CASL, but is less afflicted by the undesired magnetization transfer effects characteristic of continuous ASL. More recently a hybrid pulse-continous CASL variant (PCASL) is also used.
In a typical ASL image time-series, subtraction of the tag from the preceeding control image results in an image with intensity proportional to CBF, but does not automatically provide an absolute (quantitative) CBF value (aCBF) in units of [ml/100 g-tissue/min]. Moreover, while ASL-based CBF measurements in the grey matter are reasonably robust, white matter CBF measurements using ASL are considered to be not very reliable (also due to substantially lower CNR).
The aCBF at voxel "v" can be computed based on the General Kinetic Model (GKM), which involves the following components: the "delivery" function (i. e. the normalized arterial concentration of magnetization arriving at the voxel "v"), the "residue" function (i. e. the fraction of tagged water) and the "magnetization relaxation" function (i. e. the fraction of the original longitudinal magnetization tag carried by the blood water). The so valled Standard Kinetic Model (SKM) is a special case of the GKM, and is widely used for absolute CBF calculation. The SKM makes the important assumptions that (i) no tag arrives before the so called "transit delay" (or after the "transit delay" + "tag duration") period, (ii) the blood-tissue water exchange follows a single-exponential model, and (iii) the blood water is completely exchanged with tissue water after arrival at the voxel "v", thereby the magnetization continues to decay at the T1 of tissue.
In the PASL case the SKM can be expressed as:
aCBF[v] = (dM[v] * lambda) / ( 2 * alpha * MoA[v] * TI1 * exp( -TI2[slice] / T1A) )
In the CASL case:
aCBF[v] = (dM[v] * lambda) / ( 2 * alpha * MoA[v] * T1A * ( exp(-w[slice]/T1A) - exp(-(tau+w[slice])/T1A) ) );
where:
dM(v) = average control - tag value for voxel v, MoA(v) = the equilibrium arterial blood magnetization
TI1 = time of the QUIPSS saturation pulse, TI2(slice) = TI2 + slice_number * slice_delay (slice_delay = the time taken to acquire each slice)
tau = tag width, w[slice] = w + slice_number * slice_delay (w: post-labelling delay, tag arrival time for the slice which contains voxel v)
T1A = T1 of arterial blood, T2sA - T2* of arterial blood
alpha = labeling efficiency, lambda = water partition
Some typical parameter values at 3 T are:
T1A = 1664 ms (T1 of arterial blood)
T1_gm = 1249 ms (T1 of grey matter)
T2sA = 106 ms (T2* of artierial blood)
T2sCSF = 75 ms (T2* of CSF [Cavusoglu, 2009])
T2s_gm = 44.2 ms (T2* of grey matter)
T2s_wm = 44.7 ms (T2* of white matter)
T2sT = 44.5 ms (average tissue T2* -- grey and white matter)
lamba_gm = 0.98 (water partition coefficient between blood and grey matter)
lambda_wm = 0.84 (water partition coefficient between blood and white matter)
lambda = 0.91 (average water partition coefficient, between blood and grey-white matter)
alpha = 0.95 (average labeling efficiency)
Other sequence parameters:
TE = echo time (set to the absolute minimum to reduce T2-type contamination)
TR = repetition time (longer TR maximizes SNR and ensures clearance of the label, e.g. in cases of large label widths or slow flow)
Buxton RB et al. A general kinetic model for quantitative perfusion imaging with arterial spin labeling. Magn Reson Med 1998; 40: 383-396 .
Wang JJ et al. Comparison of Quantitative perfusion imaging using arterial spin labeling at 1.5 and 4.0 Tesla. Magn Reson Med 2002; 48: 242-254.
Wang J et al. Arterial transit time imaging with flow encoding arterial spin tagging (FEAST). Magn Reson Med. 2003 Sep;50(3):599-607.
Williams DS et al. Magnetic resonance imaging of perfusion using spin inversion of arterial water. Proc. Natl. Acad. Sci. USA, 1992. 89: p. 212-216.
Wong EC et al. Implementation of quantitative perfusion imaging techniques for functional brain mapping using pulsed arterial spin labeling. NMR Biomed, 1997.10(4-5): p. 237-49.
Cavusoglu M et al. Comparison of pulsed arterial spin labeling encoding schemes and absolute perfusion quantification. Magn Reson Imaging 2009; 27: 1039-45.
Dai W et al. Continuous flow-driven inversion for arterial spin labeling using pulsed radio frequency and gradient fields. Magn Reson Imaging 2008; 60: 1488-97.