Skip to content

segmentation_fastsurfer

fastsurfer

Perform whole brain segmentation and surface reconstruction using deep-learning through FastSurfer.

Keywords : segmentation, fastsurfer, deepmi, cnn, deep-learning


Format : tuple val(meta), path(anat), path(fs_license)

TypeDescriptionMandatoryPattern
metamapGroovy Map containing sample information e.g. [ id:'sample1', single_end:false ]
True
anatfileAn anatomy (ex, .nii.gz).True*.{nii,nii.gz,mgz}
fs_licensefileThe path to your FreeSurfer license. To get one, go to https://surfer.nmr.mgh.harvard.edu/registration.html. Optional. If you have already set your license as prescribed by Freesurfer (copied to a .license file in your $FREESURFER_HOME), this is not required.
True*.txt

Format : tuple val(meta), path(*_fastsurfer)

TypeDescriptionMandatoryPattern
metamapGroovy Map containing sample information e.g. [ id:'sample1', single_end:false ]
True
*_fastsurferdirectoryFastsurfer segmentation outputs.True*_fastsurfer

Format : path(versions.yml)

TypeDescriptionMandatoryPattern
versions.ymlfileFile containing software versionsTrueversions.yml

TypeDescriptionDefaultChoices
acq3TbooleanUse this flag if the input T1w image was acquired at 3T.False
cerebnetbooleanUse this flag to enable cerebellum sub-segmentation with CerebNet.False
hypvinnbooleanUse this flag to enable hypothalamus sub-segmentation with HypVinn.False
seg_onlybooleanUse this flag to only run the segmentation part of FastSurfer. If false, will also run the surface reconstruction part (recon-surf).
True
single_threadbooleanIf true, the command will be run in single-threaded mode. By default, the command will use multiple threads based on the number of CPUs allocated to the task.False

DescriptionDOI
FastsurferThe FastSurfer pipeline consists of two main parts for segmentation and surface reconstruction.


Last updated : 2026-05-12