Obtain masks from selected labels of a segmentation volume.
In other words, for input images containing multiple labels e.g. 1=GM, 2=GM,
nums2mask
can create output images containing single labels like an image containing WM only.
apptainer exec docker://fnndsc/pl-nums2mask nums2mask [-p PATTERN] [-m MASK_SPEC] input/ output/
input/
is a directory which contains segmentations as MINC files.
--mask
is a required parameter of nums2mask
. It is one of:
- a CSV string of integers
- a whitespace-separated list of key-value pairs, separated by
:
, values being CSV-string of integers
You have a directory of input MINC segmentation files where white matter (WM) is labeled by 160 and 5 on the left, and you want to extract left WM masks.
apptainer exec docker://fndsc/pl-nums2mask nums2mask -m '160,4' input/ output/
On an input dataset containing input/1.mnc input/2.mnc
, the outputs will be
input/1_mask.mnc input/2_mask.mnc
.
You have a directory of input MINC segmentation files where white matter (WM) is labeled by 161,5 on the left, and 160,4 on the right*. You want to extract left and right WM masks to the file names "lh.wm.mnc" and "rh.wm.mnc" respectively:
apptainer exec docker://fndsc/pl-nums2mask nums2mask -m 'lh.wm.mnc:161,5 rh.wm.mnc:160,4' input/ output/
On an input dataset containing input/1.mnc input/2.mnc
, the outputs will be
output/1/lh.wm.mnc output/1/rh.wm.mnc output/2/lh.wm.mnc output/2/rh.wm.mnc
.
*Footnote: example label numbers come from FreeSurfer's
FreeSurferColorLUT
.
docker build -t localhost/fnndsc/pl-nums2mask .
docker run --rm localhost/fnndsc/pl-nums2mask chris_plugin_info > nums2mask.json