This presentation will describe extensions of the standard ASL method which enable quantification of extra haemodynamic parameters, additional to tissue perfusion. These parameters include arterial and tissue arrival times, water exchange times between the intra- and extra-vascular compartments, and estimation of partial volume effects. Changes to the acquisition scheme required to achieve these extra measurements, using multiple inflow times and echo times, will be described, and examples of uses of the techniques in vivo will be shown.
Research scientists and clinicians interested in learning more about arterial spin labeling (ASL), and the information that advanced implementations of the technique can provide beyond just tissue perfusion.
Arterial Spin Labelling is a non-invasive MR technique for measuring tissue- level blood flow, or perfusion [1, 2]. By modifying the way in which ASL data is acquired, it is possible to obtain more information from the images than just tissue perfusion. In this talk, I will describe modifications to standard ASL acquisition schemes which enable extra haemodynamic and physiological parameters to be estimated, and provide the theoretical basis for these extensions of ASL. With this knowledge, researchers and clinical scientists will be in a position to design ASL protocols which are best suited to their applications, and which enable quantification of parameters which are important and relevant to their particular study or patient group.
Traditionally, ASL images are acquired with a single inflow time, TI (or post-labelling delay, PLD, as it is usually called in pseudo-continuous ASL (pCASL)), and a single echo time, TE [1]. If the intention is to efficiently and precisely quantify perfusion, then choosing the shortest TE and a TI/PLD optimised for the best balance between SNR and insensitivity to bolus arrival times is a sensible approach [3, 4]. However, by collecting ASL data with multiple TI/PLD and/or TE values, extra useful information which is complementary to perfusion can be obtained. The aim of this presentation is to describe the practical and theoretical details of multi-TI/PLD and multi-TE ASL sequences, and explain the pros and cons of each.
Multi-TI/PLD approaches can improve the reliability of perfusion estimation by reducing assumptions about the haemodynamics of the organ being imaged, and also provide direct measurements of arterial arrival time (AAT) and tissue arrival time (TAT). The extra dynamic information available from multi-TI/PLD images can also enable the use of more advanced and complex biophysical models for parameter quantification, which expands the scope of the ASL acquisition and also has the potential to improve accuracy of the original parameters (i.e. perfusion and blood arrival times). For example, the simple single compartment model can be extended to two compartments (intra- and extra-vascular) or more [5, 6]. In the brain, the different kinetics of different tissues means that if individual voxels contain contributions from a range of tissue types, partial volume correction can be performed based on ASL data acquired over multiple TIs [7].
Several acquisition schemes exist to collect multi-TI/PLD ASL data. The most straightforward involves repeating the ASL acquisition several times with different TI/PLD values [8], but this has the disadvantage of significantly extending the overall scan time, unless the number of TI/PLDs is offset by a corresponding reduction in the number of averages for each. Improvements in measurement efficiency can be achieved by choosing the TI/PLDs based on calculated optimal values [9-12], but scan times still tend to be long. Alternative approaches to avoid excessively long scan times have included the use of a Look-Locker readout [13] and modification of the pCASL pulse train using a ‘time encoded’ labeling module to enable reconstruction of several images with different effective PLD from a single acquisition [14, 15].
Multi-TI/PLD ASL has been particularly applied in cerebrovascular disease, where disruption of the arterial blood supply is clearly likely to affect the arrival time of the ASL bolus [16]. In addition, elderly subjects tend to display increased arrival times relative to younger subjects [17, 18], and it is possible that ATT and TAT may provide useful complementary information in combination with CBF in the study and early identification of neurodegenerative disease.
It is well known that the T2 value of a tissue is dependent on both its microstructural composition and its physiological state. For example, the transverse blood relaxation rate (i.e. 1/T2blood) is proportional to its fractional deoxyhaemoglobin content, which forms the basis for BOLD fMRI [19]. The T2 value of blood water in the arterial and capillary compartments is also different from that of parenchymal tissue water, and by measuring the transverse decay rate of the ASL signal it is possible to estimate the relative proportion of signal in each compartment directly. This can be done either by using a single ‘mean’ T2 per voxel [20] or by using a two compartment, biexponential decay model for blood and tissue water [21]. Changes in the relative size of the blood and tissue compartments as function of TI/PLD indicate how quickly water exchanges from the blood into the tissue, which has the potential to offer insight into the permeability of the blood:brain barrier and activity of its acquaporin-4 channels. In addition to haemodynamic timing parameters, this approach has also been used to estimate arterial blood volume [22]. Lastly, acquisition schemes using combined T2 and T2* mapping have been used with a more complex biophysical model to also investigate the dynamics of water exchange between blood and tissue [23].
Single TI/PLD ASL is currently recommended by the ISMRM Perfusion Study Group for mapping brain perfusion [1]. However, by acquiring ASL data with multiple TI/PLDs and/or multiple TEs, we can gain insight into the dynamic delivery of the labeled bolus into the tissue of interest, and measure arrival times in the vasculature and parenchymal tissue, as well as improving the accuracy of perfusion quantification. As ASL is a completely non-invasive technique, these measurements can be done in any patient population, without any of the safety concerns relating to other injectable tracers, and can be repeated as frequently as desired for follow-up and to track the effects of therapies and interventions. As scanner performance improvements lead to higher SNR images and more efficient single shot acquisition methods, advanced ASL techniques hold great promise for the future of quantitative MR imaging.
1. 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. Magn Reson Med, 2015. 73(1): p. 102-16.
2. Gunther, M., Perfusion imaging. J Magn Reson Imaging, 2014. 40(2): p. 269-79.
3. Alsop, D.C. and J.A. Detre, Reduced transit-time sensitivity in noninvasive magnetic resonance imaging of human cerebral blood flow. Journal Of Cerebral Blood Flow And Metabolism, 1996. 16: p. 1236-1249.
4. Wong, E.C., R.B. Buxton, and L.R. Frank, Quantitative imaging of perfusion using a single subtraction (QUIPSS and QUIPSS II). Magnetic Resonance In Medicine, 1998. 39: p. 702-708.
5. Parkes, L.M. and P.S. Tofts, Improved accuracy of human cerebral blood perfusion measurements using arterial spin labeling: accounting for capillary water permeability. Magnetic Resonance In Medicine, 2002. 48: p. 27-41.
6. Li, K.L., et al., Four-phase single-capillary stepwise model for kinetics in arterial spin labeling MRI. Magn Reson.Med., 2005. 53(3): p. 511-518.
7. Chappell, M.A., et al., Partial volume correction of multiple inversion time arterial spin labeling MRI data. Magn Reson.Med., 2011. 65(4): p. 1173-1183.
8. Chalela, J.A., et al., Magnetic resonance perfusion imaging in acute ischemic stroke using continuous arterial spin labeling. Stroke, 2000. 31: p. 680-687.
9. Xie, J., et al., Optimal design of pulsed arterial spin labeling MRI experiments. Magn Reson.Med., 2008. 59(4): p. 826-834.
10. Xie, J., et al., Real-time adaptive sequential design for optimal acquisition of arterial spin labeling MRI data. Magn Reson Med, 2010. 64(1): p. 203-10.
11. Owen, D., et al., Optimisation of Arterial Spin Labelling Using Bayesian Experimental Design, in Medical Image Computing and Computer-Assisted Intervention -- MICCAI 2016: 19th International Conference, Athens, Greece, October 17-21, 2016, Proceedings, Part III, S. Ourselin, et al., Editors. 2016, Springer International Publishing: Cham. p. 511-518.
12. Kramme, J., et al., Improving perfusion quantification in arterial spin labeling for delayed arrival times by using optimized acquisition schemes. Z Med Phys, 2015. 25(3): p. 221-9.
13. Petersen, E.T., T. Lim, and X. Golay, Model-free arterial spin labeling quantification approach for perfusion MRI. Magn Reson.Med., 2006. 55(2): p. 219-232.
14. Wells, J.A., et al., In vivo hadamard encoded continuous arterial spin labeling (H-CASL). Magn Reson Med, 2010. 63(4): p. 1111-1118.
15. von Samson-Himmelstjerna, F., et al., Walsh-Ordered Hadamard Time-Encoded Pseudocontinuous ASL (WH pCASL). Magnetic Resonance in Medicine, 2016. 76(6): p. 1814-1824.
16. Zaharchuk, G., et al., Comparison of arterial spin labeling and bolus perfusion-weighted imaging for detecting mismatch in acute stroke. Stroke, 2012. 43(7): p. 1843-8.
17. Parkes, L.M., et al., Normal cerebral perfusion measurements using arterial spin labeling: reproducibility, stability, and age and gender effects. Magnetic Resonance In Medicine, 2004. 51: p. 736-743.
18. Campbell, A.M. and C. Beaulieu, Pulsed arterial spin labeling parameter optimization for an elderly population. J.Magn Reson.Imaging, 2006. 23(3): p. 398-403.
19. Thulborn, K.R., et al., Oxygenation dependence of the transverse relaxation time of water protons in whole blood at high field. Biochim.Biophys.Acta., 1982. 714: p. 265-270.
20. Wells, J.A., et al., Characterizing the origin of the arterial spin labelling signal in MRI using a multiecho acquisition approach. J.Cereb.Blood Flow Metab, 2009.
21. Wells, J.A., et al., Measuring biexponential transverse relaxation of the ASL signal at 9.4 T to estimate arterial oxygen saturation and the time of exchange of labeled blood water into cortical brain tissue. J Cereb Blood Flow Metab, 2013. 33(2): p. 215-24.
22. Liu, P., J. Uh, and H. Lu, Determination of spin compartment in arterial spin labeling MRI. Magn Reson.Med., 2011. 65(1): p. 120-127.
23. He, X., M.E. Raichle, and D.A. Yablonskiy, Transmembrane dynamics of water exchange in human brain. Magn Reson Med, 2012. 67(2): p. 562-71.