The present study assesses the effects of cross-term interactions between diffusion and imaging gradients in magnetic resonance imaging q-space analysis, and corrects for those effects for both spin echo and stimulated echo diffusion-weighted sequences. These corrections are demonstrated experimentally in unrestricted media for water and theoretically by simulating the case of restricted diffusion in a sphere. By correcting for the cross-term interactions, large imaging gradients can be used without compromising the results. Ignoring cross-term interactions could lead to a misunderstanding of the q-space analysis; for instance, the microstructural size of the sample could be overestimated, or isotropic media could be misinterpreted as being anisotropic.
Copyright 2004 Wiley-Liss, Inc.