Angular smoothing and radial regularization of ODF fields: application on deterministic crossing fiber tractography

The advent of high angular resolution diffusion imaging (HARDI) has opened up new perspectives for the delineation of crossing and branching fiber pathways. However, image acquisition under clinical conditions with limited measurement time faces the problem of poor spatial and angular resolution and the technique's high susceptibility to noise. In this paper we present a straightforward spatial filter for ODF fields that uses the data-inherent structural information around a voxel as part of a directionally selective method for angular smoothing and radial regularization (ASRR). Especially in regions where fibers cross (multimodal voxels), the method allows us to reduce noise, improve the accuracy of ODF diffusion peaks, and strengthen signals of non-dominant fibers. Moreover, we propose a dynamic scheme in which regularization is applied only to ODFs classified as multimodal. The approach is quantitatively evaluated on synthetic datasets of various configurations. With an in vivo dataset of a human subject, measured under clinical imaging conditions, we demonstrate the method's ability to improve tractography of non-dominant transcallosal fiber pathways and the long fibers of the superior longitudinal fasciculus.