Introduction to Diffusion MRI data

  • dMRI data is represented as a 4-dimensional image (x,y,z,diffusion directional sensitivity)
  • dMRI data is sensitive to a particular direction of diffusion motion. Due to this sensitivity, each volume of the 4D image is sensitive to a particular direction

Preprocessing dMRI data

  • Many different preprocessing pipelines, dependent on how data is acquired

Local fiber orientation reconstruction

  • Provides an estimation of the local (voxel-wise) underlying fiber orientation
  • Local fiber orientation reconstruction is the primer to all dMRI derivatives

Diffusion Tensor Imaging (DTI)

  • DTI is one of the simplest and most common models used
  • Provides information to infer characteristics of axonal fibres

Constrained Spherical Deconvolution (CSD)

  • CSD uses the information along more gradient encoding directions
  • It allows to resolve complex fiber configurations, such as crossings

Tractography

  • Provides an estimation of the long range underlying fiber arrangement
  • Tractography is central to estimate and provide measures of the white matter neuroanatomy

Local tractography

  • Local tractography uses local orientation information obtained from diffusion MRI data
  • Tractography requires seeds to begin tracking and a stopping criterion for termination

Deterministic tractography

  • Deterministic tractography methods perform tracking in a predictable way

Probabilistic tractography

  • Probabilistic tractography incorporates uncertainty to the tracking process
  • Provides tractograms that explore more white matter axonal fibers