Skip to content

Latest commit

 

History

History
42 lines (21 loc) · 2.41 KB

readme.md

File metadata and controls

42 lines (21 loc) · 2.41 KB

M-CRIB atlas

Update

Individual T2-weighted, T1-weighted and parcellated images comprising the M-CRIB atlas (with Desikan-Killiany compatible cortical regions) are now available via Open Science Framework here: https://osf.io/2yx8u/

Corresponding individual data for the M-CRIB 2.0 atlas (with Desikan-Killiany-Tourville compatible cortical regions) are available here: https://osf.io/4vthr/

Please use the datasets available via OSF as they will provide more accurate labeling than the group templates provided on the current page.

Files available on the current page are group templates including:

  • T2- and T1-weighted templates
  • Hard segmentation images
  • Structure probability maps
  • Segmentation label files

T2- and T1-weighted templates are based on the 10 M-CRIB participants, constructed with ANTS using nonlinear symmetric diffeomorphic normalisation (SyN) as the transformation model. Further details are available in the paper referred to below.

Thresholded hard segmentation images were constructed by registering the 10 parcellations to the T2-weighted template, and selecting the most frequently occurring label across the sample at each voxel. These segmentations have been thresholded in increments of 0.1 probability. Thus, for an image thresholded at 0.7, the given label in any voxel was present in at least 7 of the 10 participants. Sub-threshold voxels have a value of 0. For (effectively) unthresholded hard segmentation, please use the image labeled '0.1'.

Individual structure probability maps are based on registration of binarised labels for each structure for the 10 participants to the T2-weighted M-CRIB template.

Label files in ITK and FreeSurfer formats list structure indices, names, RGB and alpha values.

Templates and hard segmentations based on rigid and affine registration are also available, purely for the purpose of further illustrating the variability in morphology in the M-CRIB sample. However, these will produce suboptimal labeling, so are not recommended for this purpose. These are available in the current repository under the branch 'linear_atlas_versions'. To access this,

git checkout linear_atlas_versions

Images are in the subfolder linear_M-CRIB_atlas_versions

If available, we may add further M-CRIB atlas data via this page.

Please cite this paper when using and referring to the M-CRIB atlas.