Using the USCBrain and other atlases with BrainSuite
The following description is for USCBrain atlas but it also works for other SVReg atlases. Download and unzip the USCBrain atlas, move the new atlas folder to
The USCBrain atlas is fully compatible with BrainSuite. It can also be used separately with SVReg module in BrainSuite from the command line. Please refer to Usage.
Using the USCBrain atlas with BrainSuite GUI
- Once you download and unzip the USCBrain atlas move the new atlas folder to
/svreg folder and keep it along with other atlases there. e.g. On Windows: C:\Program Files\BrainSuite17a\svreg\
- Edit the SVReg manifest file, svregmanifest.xml, in the
/svreg directory to add your new atlas. This will enable the BrainSuite GUI to find it. The file currently contains entries for two atlases. You can add USCBrain by editing it as follows:
- GUI Usage: The next time the BrainSuite GUI is loaded, the new atlas should appear in the SVReg dialog box.
Using the USCBrain atlas with FreeSurfer
The USCBrain atlas can be used with FreeSurfer for surface labeling of a given subject using scripts for using USCBrain with freesurfer. Download and unzip the linked code. Change the working directory to the unzipped folder and do the followings steps.
- Process the subjct data using freesurfer recon-all pipeline.
- Run the following python code.
python freesurfer_label_USCBrain.py -a path-to-USCBrain-dir path-to-freesurfer-sub-dir
Where path-to-freesurfer-sub-dir is FreeSurfer processed subject directory.
The output of the registration is stored as two dfs files, one per hemisphere, inside the surf directory in freesurfer subject directory. It can be visualized in BrainSuite.
Using the USCBrain atlas with FSL
flirt -ref anat-image-BFC-file -in uscbrain-BFC-file -omat out-mat-file
fnirt --ref=anat-image-BFC-file --in=uscbrain-BFC-file --aff=out-mat-file --cout=fnirtcoeff-coeff-file
applywarp -r anat-image-BFC-file -i uscbrain-BFC-file -o warped-labels -w fnirtcoeff-coeff-file
Note that execution time for flirt is 5-10 min and fnirt can take upto 1-2 hours.