FAST
From hpcwiki
FAST, FMRIB's Automatic Segmentation Tool, is an application that make a tissue classidication and bias filed correction. There are two ways of run this tool:
GUI | Command - Line |
---|---|
[~~]$ Fast
|
[~~]$ fast
|
Fast GUI
- Number of input channels
- Path to the structural skull removed image (brain result of Bet process)
- Kind of the input image
- Path to the resulting image basename (for all the resulting images)
- Number of classes in the k-means process
- Output images
- Binary segmentation (Also put one image per class)
- Partial Volume maps
- Restored input
- Estimated bias field
- Advanced options
- Main MRF parameter
- Number of iteration of bias field removal
- Bias field smoothing
- Use a-priori probability maps for initialization
- Standard to input FLIRT transform
- Use file of initial tissue type means
The simple way to run the application is by setting the input image, the result of Brain Extraction Tool, and execute. Then segmentation process start, it takes 15 minutes approximately to classify the brain in three classes, in this case White Matter (WM), Gray Matter (GM) and Cerebro Spinal Fluid (CSF). Once the process is complete it generates the following results:
-
structural_brain_mixeltype.nii.gz
An image with -
structural_brain_pve_0.nii.gz
An image with the first class segmented. -
structural_brain_pve_1.nii.gz
An image with the second class segmented. -
structural_brain_pve_2.nii.gz
An image with the third class segmented. -
structural_brain_pveseg.nii.gz
An image with -
structural_brain_seg.nii.gz
An image with
Note that one image is created for each class. In this case, three classes (WM, GM and CSF) were consider and three images were created (pve_0, pve_1, pve_2).
To see the results run the following command:
fslview structural_brain_pve_0 -l Green -b 0.5,1 \
structural_brain_pve_1 -l Blue-Lightblue -b 0.5,1 \
structural_brain_pve_2 -l Red-Yellow -b 0.5,1 &
The image shows the segmentation results, were pve_0 corresponds to CSF, pve_1 corresponds to GM, and pve_2 corresponds to WM
Fast Command line
To run the application for the command line it is necessary to consider:
[~~]$ fast [options] file(s)
Optional arguments (You may optionally specify one or more of):
-n,--class number of tissue-type classes; default=3
-I,--iter number of main-loop iterations during bias-field removal; default=4
-l,--lowpass bias field smoothing extent (FWHM) in mm; default=20
-t,--type type of image 1=T1, 2=T2, 3=PD; default=T1
-f,--fHard initial segmentation spatial smoothness (during bias field estimation);
default=0.02
-g,--segments outputs a separate binary image for each tissue type
-a <standard2input.mat> initialise using priors; you must supply a FLIRT transform
-A <prior1> <prior2> <prior3> alternative prior images
--nopve turn off PVE (partial volume estimation)
-b output estimated bias field
-B output bias-corrected image
-N,--nobias do not remove bias field
-S,--channels number of input images (channels); default 1
-o,--out output basename
-P,--Prior use priors throughout; you must also set the -a option
-W,--init number of segmentation-initialisation iterations; default=15
-R,--mixel spatial smoothness for mixeltype; default=0.3
-O,--fixed number of main-loop iterations after bias-field removal; default=4
-H,--Hyper segmentation spatial smoothness; default=0.1
-v,--verbose switch on diagnostic messages
-h,--help display this message
-s,--manualseg <filename> Filename containing intensities
-p outputs individual probability maps fsl5.0-fast
To execute the classification with the default options, command line instruction is:
[~~]$ fast -o structural_braincl structural_brain
One more time to review the results, open the images in FSLView
fslview structural_braincl_pve_0 -l Green -b 0.5,1 \
structural_braincl_pve_1 -l Blue-Lightblue -b 0.5,1 \
structural_braincl_pve_2 -l Red-Yellow -b 0.5,1 &
As in the GUI app, the results correspond to the parameters selected for the process, in this case the default options.