Differences between revisions 11 and 12
Deletions are marked like this. Additions are marked like this.
Line 92: Line 92:

Cortical Hubs

The 'Cortical Hubs' (CHubs) are regions-of-interest defined by cortical thickness differences between older controls and individuals classified as having AD, created using the OASIS data set. These ROIs were used in the study The Dynamics of Cortical and Hippocampal Atrophy in Alzheimer Disease, Sabuncu et al.. See also: Supplemental, and below is a snapshot of a figure from the paper.

Freesurfer v5.2 (and v5.1 local) contains these ROIs as labels, found in the directory $FREESURFER_HOME/subjects/fsaverage/label. There are eight:

FS label name

name used in paper


inferior frontal cortex


inferior parietal cortex


inferior parietal sulcus


lateral temporal cortex


not used in paper


entorhinal cortex (medial temporal lobe)


posterior cingulate


temporalpolar cortex

These labels are also combined into .annot files: ?h.oasis.chubs.annot

To view these on the fsaverage rh surface:

tksurfer fsaverage rh inflated -annotation oasis.chubs

To map these labels to your subject set (which you would define in a SUBJECTS csh shell var), first you will need to symlink the fsaverage subject into your own subjects dir:

ln -s $FREESURFER_HOME/subjects/fsaverage

Confirm that you see 16 .label files and two .annot files:

ls $SUBJECTS_DIR/fsaverage/label/*chubs*

Now you can do the mapping:

foreach s ($SUBJECTS)
    echo $s
    foreach h (lh rh)
        foreach l (\
  .oasis.chubs.ifc \
  .oasis.chubs.ipc \
  .oasis.chubs.ips \
  .oasis.chubs.lateraltemporal \
  .oasis.chubs.mtl \
  .oasis.chubs.retrosplenial \
            set cmd=(mri_label2label \
                --srclabel fsaverage/label/${h}${l} \
                --srcsubject fsaverage \
                --trgsubject $s \
                --trglabel $s/label/${h}${l} \
                --regmethod surface \
                --hemi $h)
            echo $cmd

You will probably want to have these labels combined into a .annot file so that mris_anatomical_stats can operate on this set of labels, so to create this .annot file, first you will need a colortable. Create this text file, naming it colortable_oasis_chubs.txt, and put it in your $SUBJECTS_DIR/scripts dir:

  0  unknown                        25   5  25   0
  1  oasis.chubs.ifc                25 100  40   0
  2  oasis.chubs.ipc               125 100 160   0
  3  oasis.chubs.ips               220  60 220   0
  4  oasis.chubs.lateraltemporal   120  70  50   0
  5  oasis.chubs.medialpfc         220  20 100   0
  6  oasis.chubs.mtl               220  20  10   0
  7  oasis.chubs.retrosplenial     140  20 140   0
  8  oasis.chubs.tp                200  35  75   0

Now you can run this script to create the file ?h.oasis.chubs.annot file in the /label directory of each of your subjects:

foreach subj ($SUBJECTS)
  foreach hemi (lh rh)
    set cmd=(mris_label2annot \
      --s $subj
      --hemi $hemi \
      --ctab $SUBJECTS_DIR/scripts/colortable_oasis_chubs.txt \
      --l $hemi.oasis.chubs.ifc.label \
      --l $hemi.oasis.chubs.ipc.label \
      --l $hemi.oasis.chubs.ips.label \
      --l $hemi.oasis.chubs.lateraltemporal.label \
      --l $hemi.oasis.chubs.medialpfc.label \
      --l $hemi.oasis.chubs.mtl.label \
      --l $hemi.oasis.chubs.retrosplenial.label \
      --l $hemi.oasis.chubs.tp.label \
      --a oasis.chubs

Last but not least, a snaphot from the Sabuncu paper: chubs.jpg


  • While the paper says these are from the OASIS data set, they are actually from an earlier data set created at Washington University whose subjects were later included in the OASIS data set. The data set used to create these CHubs is found in the NMR Center here:

  • The CHubs ROIs were mapped to the ADNI data set. The ADNI data set is found here:


These are a bunch of scripts (in the /scripts dir) which performed those mappings... see *chubs* files.

Chubs (last edited 2011-10-16 17:57:41 by NickSchmansky)