Estimation of Integral Curves from HighAngular Resolution Diffusion Imaging (HARDI) Data

Sakhanenko L,Delaura M,Zhu Dc

doi:10.15761/ndt.1000126

Abstract

Estimation of Integral Curves from HighAngular Resolution Diffusion Imaging (HARDI) Data

Highlights

We discuss the superiority of the high order tensor approach to High Angular Resolution Diffusion Imaging (HARDI), or High Angular Resolution Diffusion Tensor Imaging, over the probabilistic tractography approaches and the lower order tensor model (DTI, or Diffusion Tensor Imaging) discussed in
Since the detailed anatomy of the brain is still uncharted in terms of connectivity due to high amount of noise, this approach runs a high risk for misplacement of curve estimates and so called “phantom images”
We provide a tracing of the fiber along with surrounding confidence ellipsoids so that scientists can better understand where the true fiber is located

Summary

Introduction

We discuss the superiority of the high order tensor approach to HARDI, or High Angular Resolution Diffusion Tensor Imaging, over the probabilistic tractography approaches and the lower order tensor model (DTI, or Diffusion Tensor Imaging) discussed in. We move towards a semi-parametric high order tensor model approach of the estimation of trajectories. The methods in [1,2,3] for DTI and HARDI models, both semiparametric approaches, have this advantage.

Results

Conclusion