Fiber Tracking and DTI

Question 1

What is the structure and function of white matter?

Answer 1

• consists of fibers
• connects different parts of the cortex and subcortical areas

Question 2

What is the corpus callosum?

Answer 2

white matter connecting both hemispheres

Question 3

What is diffusion?

Answer 3

natural tendency of particles to distribute evenly in a soluble solution
- based on Brownian motion

Question 4

How does diffusion allow to track white matter fibers in the brain?

Answer 4

• diffusion of water is restricted by fat: water does not diffuse through fat but along it
• neurons are insulated by oligodendrocytes
• insulation happens through myelin sheets basically consisting of fat
• myelin follows the orientation of the axon bundle / fiber and water will diffuse parallel (and never perpendicular) to these

Question 5

What is the difference between isotropic and anisotropic diffusion?

Answer 5

• isotropic: random diffusion in all directions, no restrictions, e.g. in ventricles
• anisotropic: strength of diffusion varies for different directions, e.g. corpus callosum

Question 6

What mathematical structure is used to describe diffusion?

Answer 6

• tensor
• described by 9 parameters
• isotropic diffusion: ball-shaped
• anisotropic diffusion: cigarette-shaped
• using matrix operations from dimensionality reduction, principial eigenvectors of the tensor are obtained
• lengths and orientations of eigenvectors show how much diffusion happens in every direction

Question 7

What is fractional anisotropy and what is it used for?

Answer 7

• summary statistic (scalar) collapsing the 9 values describing a tensor
• scales from 0 (isotropic) to 1 (most anisotropic)
• used to weigh the connection between 2 brain areas by calculating the mean FA of all voxel of the fiber tract connecting the 2 regions (measure of strength of connection)
• length of connection: number of voxels of the streamline

Question 8

Fiber assignment by continuous tracking (FACT)

Answer 8

• place seed-voxel in every single voxel of the brain
• start with one voxel in fiber tract and follow a streamline going the direction of the tensor / diffusion
• stop criteria (e.g. turn too sharp)
• not done in voxel space but in a vector space with way better resolution

Question 9

individual differences in connectivity

Answer 9

• only 7% of possible connections present in 215 subjects
• 20% of possible connections not present in any participant

Question 10

features of brain connectivity

Answer 10

• higher connectivity within hemispheres than between them
• higher connectivity within subcortical regions than between them
• subcortical regions preferably connect to other brain areas in same hemisphere
• exponential distribution of nodal degree/strength: few areas are very well connected (hubs), others are sparsely connected
• short communication paths and high local clustering
• brain networks are modular: groups of nodes that are strongly interconnected, but weakly connected to nodes outside the network

Question 11

The rich club

Answer 11

• brain networks with most connections (hubs) preferably connect to each other and stronger so than you would expect from the sheer number of connections
• distant brain regions are almost always connected through the rich club

Question 12

Empirical findings regarding the rich club

Answer 12

• without rich club connections, communicability drops drastically
• stronger connectivity within the rich club correlates with better performance in general cognition and executive control tasks
• difference in rich-club connections between schizophrenia patients and healthy controls

Question 13

Which brain network has the highest correspondence between functional and structural connectivity

Answer 13

default mode network