Characterizing diffusion tensor imaging data with directional entropy

We describe the use of directional entropy (DE) in the directional analysis of diffusion tensor imaging (DTI) data. The directional entropy is a measure of disorder in a directional distribution. It could provide a relatively simple, yet meaningful measure about the brain white matter integrity, complementary to the traditional measures used, such as mean diffusivity or indices of diffusion anisotropy. The challenge of the DTI is to produce measures that would be easily comparable across subject and patient populations. We studied directional distributions and entropy with simulations and measured DTI data. Directional entropy could serve as an additional measure to characterize developmental or pathological states in brain