The effect of gradient sampling schemes on measures derived from diffusion tensor MRI: A Monte Carlo study†

There are conflicting opinions in the literature as to whether it is more beneficial to use a large number of gradient sampling orientations in diffusion tensor MRI (DT-MRI) experiments than to use a smaller number of carefully chosen orientations. In this study, Monte Carlo simulations were used to study the effect of using different gradient sampling schemes on estimates of tensor-derived quantities assuming a b-value of 1000 smm(-2). The study focused in particular on the effect that the number of unique gradient orientations has on uncertainty in estimates of tensor-orientation, and on estimates of the trace and anisotropy of the diffusion tensor. The results challenge the recently proposed notion that a set of six icosahedrally-arranged orientations is optimal for DT-MRI. It is shown that at least 20 unique sampling orientations are necessary for a robust estimation of anisotropy, whereas at least 30 unique sampling orientations are required for a robust estimation of tensor-orientation and mean diffusivity. Finally, the performance of sampling schemes that use low numbers of sampling orientations, but make efficient use of available gradient power, are compared to less efficient schemes with larger numbers of sampling orientations, and the relevant scenarios in which each type of scheme should be used are discussed.