Closed mahirtazwar closed 1 year ago
Set the intent_code
of your NIfTI header to 1002 to specify NIFTI_INTENT_LABEL. This will use the same random color scheme as FSLeyes and use nearest neighbor interpolation (to solve the partial volume issue that a voxel that is a mixture of area 17 and area 19 should not be identified as area 18).
To see examples, choose any of the images listed under the File/AddAtlas menu item.
How to create a random color look up table so that each label in a segmentation image gets a random color instead of a gradient (similar to the random colormap in fsleyes, or how segmentations are loaded on ITK-SNAP)? In this case, each region of interest with a unique label will have a unique or repeated color (since I have more than 200+ ROIs).
Reference to: _Originally posted by @mahirtazwar in https://github.com/neurolabusc/MRIcroGL10_OLD/issues/35#issuecomment-1626337859_