Freesurfer Guide
Running Freesurfer on the BIAC cluster
An easy way to run FreeSurfer is to submit a cluster batch job using the QSUB command with a standard FreeSurfer batch processing script. To do so, you need just 2 things:
- a high resolution whole-brain T1-weighted anatomical DICOM image data set within a BIAC experiment directory
- a name you want to assign for the FreeSurfer reconstruction of this subject
The standard FreeSurfer batch processing script is located on the cluster head node at:
/usr/local/packages/qsub_templates/RunFreeSurf.sh
You can use this script with no modification by entering the following command on the cluster head-node (not on a qinteract node):
qsub -v EXPERIMENT=MyExperiment /usr/local/packages/qsub_templates/RunFreeSurf.sh SubjID SeriesDir
where:
- MyExperiment is your BIAC experiment name (eg. Expt.02)
- SubjID is the name you want to assign for this subject reconstruction (eg. 1234 or subj21)
- SeriesDir is the directory path to get to your anatomical DICOM images for the subject you want to reconstruct, within your MyExperiment directory (eg. Data/Anat/20110101_12345/series300 )
The job will create an output directory for this subject, within your MyExperiment directory at:
Analysis/freesurfer/SubjID
Example:
To perform a full FreeSurfer reconstruction of anatomical series 300 of scanner exam 10123 acquired on Feb 1, 2011 under BIAC experiment BrainScan.02, and store the results as subject BS123, enter the command:
qsub -v EXPERIMENT=BrainScan.02 /usr/local/packages/qsub_templates/RunFreeSurf.sh BS123 Data/Anat/20110201_10123/series300
The results will be found (about 20 hours later) in the following directory within the BrainScan.02 experiment directory:
Analysis/freesurfer/BS123
Notes:
The QSUB command will automatically insert your email address to notify you of job
status.
It takes about 20 hours for a full FreeSurfer reconstruction to run on the cluster.
See Also:
Another sample template script (requires some editting): FreeSurfer template
Obsolete: Freesurfer on Golgi