This study aims to better understand the information provided by diffusion weighted magnetic resonance imaging (MRI) in tissues. As extensive uses of diffusion-weighted imaging (DWI) techniques evolve, it is essential to develop a greater understanding of the factors that affect diffusion in tissues. The specific issues we will address are: (l) Quantitative determination of the magnitude and time course of changes in water diffusion coefficient that happen during pathological occurrences such as stroke and seizure, using animal models we have developed. Diffusion- weighted imaging will be obtained with time resolution on order of seconds, and during the acute stages of stroke and seizure; (2) Quantitative evaluation of different mechanisms responsible for the alteration of apparent water diffusion coefficient (ADC), including changes in restriction of water diffusion during pathological changes, cytosolic streaming motion, and variations of local magnetic field gradient due to susceptibility difference caused by oxyhemoglobin/ deoxyhemoglobin conversion. We will use specifically designed experiments in simple phantoms, perfused cells, freshly excised tissues, and animal models to address each of these; (3) Develop a model for water diffusion in heterogeneous systems such as tissues based on a clear understanding and quantitative evaluation of each individual mechanism that affects water diffusion. Our own preliminary observation of reduction in ADC during seizure has highlighted the need to quantitatively validate the hypotheses concerning ADC reduction in ischemia that have been suggested by other researchers. Since ischemia and seizure represent two quite different biological conditions (blood flow, oxygenation, and energy status, etc), close comparison of the two models and quantitative studies of individual mechanisms should facilitate improved understanding of the ADC changes in both. We will use NMR spectroscopic and imaging methods based on relaxation and diffusion measurements to quantify water transport among diffusion barriers and across cell membranes or capillary walls of finite permeability. We will use numerical analysis and computer simulations to quantify diffusion among barriers of different shapes, sizes, and different boundary conditions. We will use the NMR q-space concepts developed by Callaghan (1991) to study microstructure and dynamics beyond the resolution of conventional MRI. The q-space imaging is based on the pulsed gradient spin-echo (PGSE) method first developed by Stejeskal and Tanner (1965), and it can be used to characterize water displacement profiles which reflect, if analyzed appropriately, the autocorrelation function of compartment dimensions as well as the relative number and sizes of differently diffusing compartments. These new methods are potentially very powerful at providing new insights into diffusion in heterogeneous compartmented systems such as tissue, but to date their use has been restricted largely to inanimate samples. We will perform experiments on our 2T and 7T scanners both of which are equipped with high strength, shielded magnetic field gradients. A further significance of this work is that it would evaluate the value of the q-space imaging technique for biological samples.