Visualizing crossing probabilistic tracts

TL;DR: This work suggests a visualization based on Fiber-Stippling but sensible to multiple diffusion orientations from HARDI-based diffusion models, which implies that tract crossings may now be visualized as crossing stipples, which is an essential step towards an accurate visualization of the neuroanatomy.

Abstract: Diffusion weighted magnetic resonance imaging (dMRI) together with tractography algorithms allow to probe for principal white matter tracts in the living human brain. Specifically, probabilistic tractography quantifies the existence of physical connections to a given seed region as a 3D scalar map of confidence scores. Fiber-Stippling is a visualization for probabilistic tracts that effectively communicates the diffusion pattern, connectivity score, and anatomical context. Unfortunately, it cannot handle multiple diffusion orientations per voxel, which exist in high angular resolution diffusion imaging (HARDI) data. Such data is needed to resolve tracts in complex configurations, such as crossings. In this work, we suggest a visualization based on Fiber-Stippling but sensible to multiple diffusion orientations from HARDI-based diffusion models. With such a technique, it is now possible to visualize probabilistic tracts from HARDI-based tractography algorithms. This implies that tract crossings may now be visualized as crossing stipples, which is an essential step towards an accurate visualization of the neuroanatomy, as crossing tracts are widespread phenomena in the brain.