Double-wave vector diffusion weighting is expected to represent a possibility to acquire information on cell size and shape. In numerical simulations, a theoretical description of the double wave vector-weighted signal is tested in a situation where the approximating assumptions (infinitely short gradient pulses, infinitely long diffusion time, infinitely long or zero delay between diffusion weightings, small gradient moment) are not strictly met. It is shown how the dependence of the signal on the angle between the diffusion gradient directions changes upon varying the delay between the second and the third gradient pulse, and how the measures of pore geometry derived from signal differences between two angles depend on the duration, temporal separation, and amplitude of the gradient pulses. The results indicate that a violation of the approximation conditions due to finite gradient pulse duration and separation generally leads to an underestimation of pore size and pore eccentricity. It is also concluded from the simulations that for pore sizes on the order of 10 microm the investigated effects are in a range that is observable even with whole-body gradient systems.
