NMR Imaging of Flow and Perfusion using Hyperpolarized Nuclei: Theoretical Considerations and Application to Experimental Models

Sammanfattning: In the studies presented in this thesis, hyperpolarized tracers have been used for the study of macroscopic flow and capillary perfusion with magnetic resonance imaging. The feasibility of performing vascular studies using echo-planar imaging (EPI) and hyperpolarized 129Xe was investigated using xenon dissolved in ethanol prior to injection into a flow phantom. It was concluded that hyperpolarized 129Xe could only be expected to yield images of a sufficiently good SNR if depolarization losses and dilution effects could be avoided. Using the same nuclide, the signal equation for the spoiled gradient-echo sequence in the presence of flow was investigated using a similar set-up, and experimental data were compared with derived theoretical expressions. Relationships were also derived for the optimal flip angle at different flow distributions. Two investigations of perfusion with hyperpolarized 13C in labelled compounds were performed. One concern with hyperpolarized tracers is that they depolarize during the course of the perfusion examination. This behaviour was incorporated into the theory of bolus tracking. It was found that no modification of the theory was necessary to quantify tissue blood flow. Assuming an intravascular tracer, it was demonstrated that the blood volume and mean transit time (MTT) were underestimated if the MTT was long or the depolarization rate of the tracer was high. It was also shown, analytically and by simulations, that if the depolarization rate was known, this effect could be compensated for. An experimental investigation using hyperpolarized 13C for the study of cerebral blood flow in rat following venous injection was performed. Maps of the cerebral blood flow, cerebral blood volume and MTT were calculated and the MTT could be evaluated quantitatively (2.8 ± 0.8 s in five animals). A new method of assessing tissue blood flow, denoted bolus differentiation, was also proposed based on the fact that the magnetization of a hyperpolarized tracer can be destroyed permanently by subjecting the tracer to radio-frequency excitation. The technique allows for a tissue blood flow assessment that is insensitive to arterial delay and dispersion and these parameters can be determined as a by-product of the acquired image series. In an experimental investigation in rabbit kidneys, quantitative maps of the cortical blood flow (5.7/5.4 ± 1.6/1.3 ml/min per ml tissue) (mean ± SD, right/left kidney), arterial mean transit time (1.47/1.42 ± 0.07/0.07 s) and arterial dispersion (1.78/1.93 ± 0.40/0.42 s2) were calculated and evaluated in six animals.