The objective of this work is to develop techniques for quantitative non-invasive imaging of tissue perfusion using magnetic resonance imaging. The motivation for this project is too classes of applications: clinical use of perfusion imaging for non-invasive diagnosis and evaluation of therapy in acute stroke; and functional perfusion mapping of the brain, either for clinical applications such as pre-neurosurgical planning, or for basic studies of brain function. The perfusion imaging techniques used are based on magnetic labeling of arterial water using radiofrequency pulses. Current MRI based techniques can generate perfusion maps, but are inaccurate, not time efficient, and not amenable to multislice or 3D acquisition. Several effects comprise the accuracy of existing techniques including: misinterpretation of labeled blood that is passing through a slice as perfusion of the slice; a spatially varying transit delay between the application of the arterial tag and the arrival of tagged blood into the imaging slice; exchange of tagged water between vascular and extravascular compartments which have different relaxation parameters; and magnetization transfer (MT) effects. The first goal of this project is to devise imaging techniques that measure these effects, using combinations of gradient and radiofrequency pulses to selectively destroy signal from specific populations of protons, and controlling the characteristics of the arterial tag. Significant progress has been made in addressing some of these effects individually, but no single technique can claim to be truly quantitative. It is almost certain that a series of imaging techniques will be required to strictly control for all of the confounding effects, and that the resulting perfusion examination will be prohibitively time consuming. The second general goal of this project is to explore techniques that minimize, rather than measure, these confounding effects in the interest of creating a time efficient perfusion measurement, and compare these to the more rigorous techniques described above to quantify possible errors introduced into the perfusion measurement. The third goal is to explore methods for extension of these techniques to multislice or 3D. This extension is not straightforward as it is in most imaging techniques, because of the nature of the tagging schemes and MT effects, and will require particular attention to measured transit delays and the frequency dependence of MT effects. Finally, in applications where the quantity of interest is dynamic changes in perfusion, such as functional imaging or treatment of acute stroke, changes in blood oxygenation and transit times can confound the perfusion measurement, and separation of these effects from the perfusion measurement is a fourth goal.
| Wong, E C; Liu, T T; Luh, W M et al. (2001) T(1) and T(2) selective method for improved SNR in CSF-attenuated imaging: T(2)-FLAIR. Magn Reson Med 45:529-32 |
| Luh, W M; Wong, E C; Bandettini, P A et al. (2000) Comparison of simultaneously measured perfusion and BOLD signal increases during brain activation with T(1)-based tissue identification. Magn Reson Med 44:137-43 |
| Wong, E C; Luh, W M; Liu, T T (2000) Turbo ASL: arterial spin labeling with higher SNR and temporal resolution. Magn Reson Med 44:511-5 |
| Luh, W M; Wong, E C; Bandettini, P A et al. (1999) QUIPSS II with thin-slice TI1 periodic saturation: a method for improving accuracy of quantitative perfusion imaging using pulsed arterial spin labeling. Magn Reson Med 41:1246-54 |
