What does DTI measure?
A: DTI measures the diffusion of water molecules in brain tissue, exploiting the fact that diffusion is directionally restricted by white matter microstructure (axons, myelin).
What is the difference between DWI and DTI?
DWI = the raw diffusion-weighted images; shows diffusion contrast.
DTI = a model applied to DWI data that fits a diffusion tensor (3×3 matrix) per voxel.
Why can DTI infer white-matter architecture?
Because water diffuses more freely along axons than across them → anisotropy.
What causes signal attenuation in diffusion-weighted imaging?
After dephasing and rephasing gradients, moving (diffusing) spins do not fully rephase, causing signal loss.
Why must diffusion gradients be applied in at least 3 directions?
Because diffusion is a 3D process; at minimum, 6 directions are required to estimate the full tensor.
What is “isotropic” vs “anisotropic” diffusion?
Isotropic: equal diffusion in all directions → CSF; low FA.
Anisotropic: diffusion is directionally constrained → white matter; high FA.
What is Fractional Anisotropy (FA)?
A normalized measure of directional dependence of diffusion (0 = isotropic, 1 = highly anisotropic).
What does high FA indicate?
Coherent fiber orientation, intact white matter, more myelination.
What is tractography?
Reconstruction of white-matter fiber pathways by linking voxels with similar diffusion direction.
Q: What do red/green/blue colors mean in color FA maps?
Red = left–right
Green = anterior–posterior
Blue = inferior–superior
What is the biggest limitation of DTI?
It cannot resolve crossing, kissing, or branching fibers; tensor model assumes one dominant direction per voxel.
Why does FA drop at fiber crossings?
Because two orientations in one voxel reduce directional coherence, even if fibers are intact.
What does cerebellar tractography reveal (lecture example)?
Structural pathways between cerebellum and hypothalamus; relevance for emotion and aggression.
Strengths of DTI
- index of integrity of white matter
- cohrence of bundles and study of pathways
- higher intensity = better organization of fiber bundles/more myelin = faster transpor along the axons
Limitations of DTI
- at crosing points of two fiber bundles there is no longer one macher fiber direction
- FA values will be l ower in regions where bundles fan out
- Susceptible to eddy currents and movement-related artefacts
- no directional information on fiber tracks
Is DTI direction sensitive?
DTI is direction-sensitive, but it can only detect ONE dominant diffusion direction per voxel.
If more than one direction exists, DTI cannot separate them and the direction estimate becomes unreliable.
how does diffusion imaging work
Diffusion imaging works by applying two diffusion gradients. If water moves, spins pick up unpredictable phase shifts that cannot be fully rephased, causing signal loss. The amount of signal loss in each direction reveals how strongly water diffuses along that direction. Repeating this in many directions yields a diffusion tensor, whose shape and orientation reflect microstructural organization of white matter.
Explain the physics of DWI signal
Diffusion imaging uses two equal and opposite gradients. Spins that do not move between them rephase and give strong signal. Spins that move produce incomplete rephasing and lose signal. Therefore, signal attenuation reflects the magnitude of water diffusion.
what does DWI measure
how much signal is lost when you apply gradients in a certain direction.
If spins move along the gradient direction → large signal loss
If they move perpendicular to it → small signal loss
If they don’t move → almost no signal loss
Diffusion is direction-dependent.
what is a tensor and how is it related to diffusion gradients
Diffusion gradients applied in multiple directions cause direction-dependent signal attenuation.
A diffusion tensor is a 3×3 matrix that describes how strongly diffusion occurs along every direction in 3D space.
By fitting a tensor to these attenuation values, we obtain a 3D ellipsoid that captures the magnitude and direction of water diffusion in each voxel, enabling measures such as FA and principal diffusion direction.
