Investigating the uncertainty in multi-fiber estimation in High Angular Resolution Diffusion Imaging

In this paper, we investigated the reconstruction accuracy and information uncertainty in multi-fiber estimation to better understand the tradeoff between scanning time and angular precision in High Angular Resolution Diffusion Imaging (HARDI). Reconstruction accuracy was measured using the Kullback-Leibler divergence (sKL) on the orientation density functions (ODFs) first in simulations with varying b-values and variable additive Rician noise. ODFs were computed analytically from tensor distribution functions (TDFs) which model the HARDI signal at each point as a unit-mass probability density on the 6D manifold of symmetric positive definite tensors. Reconstruction accuracy rapidly increased with additional gradients at lower SNR. The information uncertainty was quantified by the Exponential Isotropy (EI), a TDF-derived measure of fiber integrity that exploits the full multidirectional HARDI signal. Simulations and empirical results both found that information uncertainty decreased as angular resolution increased, and plateaued at around 70~80 gradients. Furthermore, in high magnetic field (7 Tesla) HARDI, the reconstruction accuracy and information uncertainty index decreased at higher b-values.