T2 weighted MRI
- Imaging modality with longer TE and TR times
- Air and bone appear dark, while CSF, blood and edema appear bright.
- contrast and brightness are predominately determined by the T2 properties of tissue.
T1 weighted MRI
- Image modality with short Time Echo (TE) and short rotation time (TR).
- CSF appears dark
- Basic pulse sequences in MRI and demonstrates differences in the T1 relaxation times of tissues.
Main steps in structural MRI process
- scan acquisition (patient in machine)
- Data representation in binary
- Use visualization software to represent data as image
- Analysis with software
- statistical analysis
Motion artifacts
- A problem in structural and functional MRI.
- Higher motion = lower gray matter volume estimates
- Post hoc: exclude scans with high motion; new correction approaches for DTI data
Voxel-wise lesion-symptom mapping, 3 approaches:
- lesion-defined
- behavioral performance of a group of patients with a common area of injury - behavior-defined
- patients are grouped according to whether or not they show a specific behavioral deficit - mass-univariate
- does not require patients to be grouped by either lesion site or behavioral cutoff a priori
- for each voxel, patients are categorized according to whether they did or did not have a
lesion affecting that voxel
Manual volumetry
Manual
- Remains the gold standard for assessing changes in brain volume - ITK-Snap : hand-selected regions - Labor intensive!
Automated volumetry
pros:
- faster,
- more reliable,
- standardized
cons:
- Freesurfer can overestimate total hippocampal volumes
- Problems with accurately detecting boundaries between hippocampus and neighboring structures
FSL F.I.R.S.T. automated segmentation
“FMRIB’s Integrated Registration & Segmentation Tool”
- model-based segmentation of subcortical structures
- shape/appearance (intensity) models constructed from manually segmented images (training data)
Voxel-based morphometry:
- voxel-wise analysis of the local concentration of gray matter
- align images globally and compare GM likelihood at each voxel
- brain extraction 2. segmentation 3. normalization 4. smoothing
Spatial smoothing
- Why?: makes each voxel more similar to its neighbors
- Spatial normalization is not perfect, and smoothing helps accommodate inter-individual differences in local anatomy
Voxel-based morphometry Example study:
London Taxi drivers showed increase in gray matter density in posterior hippocampi with
concomitant changes to their memory profile
Cortical thickness
Measuring the highly-folded outer layer of grey matter with Free Surfer.
Shortest distance between white matter surface and pial
surfaces
→ 1-5 mm in healthy subjects
Cortical thickness estimation (FreeSurfer)
Pros: - automated, continuous, whole cortex - direct, biologically meaningful measure in millimeters Cons: - heavy post-processing (6-24 hours/scan) - dependent on classification - manual corrections often necessary - limited to (neo)cortex
