Case II: FSL FAST

FSL-FAST usually does a pretty good job to estimate gray matter. However a quick cleaning operation on the outer gray matter boundary with Segmentator can give it a solid finishing touch.

Software

Package	Tested version
FSL-FAST	5.0.9
Segmentator	1.2.0

Data:

• Brain extracted (fsl-bet) T1w divided by PDw images (3D MPRAGE, 0.7mm isotropic voxels).
• fsl-fast with 4 classes with export restored image to acquire inhomogeneity (bias field) corrected image.
• 2 classes are merged to create GM

Problem:

GM labels are quite noisy on the outer GM regions. Wispy structures can be seen:

Outer GM boundary should be fixed for the operations assuming correct GM segmentation.

Solution:

Manual editing or grow region tools can be used solve this problem however these methods can be very fiddly depending on the size of the data and the regions of interest.

This work can also be done by using Segmentator, much less laboriously:

1. Load the data to Segmentator:

segmentator /path/to/file.nii

2. Position the red circular mask in the 2D Histogram (left panel) in a way to leave out the unwanted tissues (e.g vessels and pial surface) in the image browser (right panel). Like in the figure below:

4. Save the mask (red voxels in the image browser) using Export Nifti button.

5. (optional) opening operation can improve the mask further (TODO: add gist).

6. Mask the nifti file:

fslmaths /path/to/file.nii.gz -mas /path/to/mask.nii.gz /path/to/file_masked.nii.gz

After this operation -tighter brain extraction if you wish-, outer GM regions looks more plausible.