The goal of the proposed research is to develop and validate magnetic resonance (MR) methods that can sensitively detect and characterize tissue perfusion. Preliminary data in our laboratory has demonstrated the feasibility of utilizing both contrast and non-contrast techniques. We will therefore explore three approaches: 1) In studies of the heart with Gd-DTPA, we will apply a comparmental kinetic model, similar to those used in nuclear medicine, to relate measured MR image intensity changes following injection of the agent to myocardial blood flow and the volume of distribution of the agent. The method will be validated in a canine model of normal and partially stenosed myocardium, and then used to test the feasibility of using MR measurements with Gd-DTPA and the vasodilator dipyridamole as an MR stress test for evaluating coronary artery disease. 2) In brain studies with Gd-DTPA, we will optimize techniques for measuring both the T1 relaxation effects, and the more pronounced transient susceptibility-induced signal loss on T2-weighted images. The physical basis of the latter effect will be clarified by phantom experiments and theroretical modeling. Preliminary data suggests that the susceptibility effect is correlated with cerebral blood flow (CBF), and we will determine the relation between both of these effects and CBF and cerebral blood volume (CBV) measured with radioactive tracers. Animal experiments will be performed in a rat model of global physiological variation and a cat model of middle cerebral artery occlusion. 3) Methods for brain imaging based on the intrinsic effects of motion on the MR signal, which require no contrast agent, will be developed to provide accurate measures of diffusion and restricted diffusion, and to separate the effects of diffusion and perfusion with a perfusion sensitive pulse sequence. These methods will be tested in phantoms and in the same animal models as in the brain Gd-DTPA experiments. The brain imaging methods will be validated in human subjects in stroke, neoplasia and degenerative dementia by direct comparison with positron emission tomography (PET) maps of CBF and CBV. Sensitivity and specificity of these methods for lesion detection and characterization will also be compared with that of conventional T1- and T2-weighted imaging.
Showing the most recent 10 out of 65 publications
