/Caret:Atlases/Conte69 Atlas

From Van Essen Lab

(Difference between revisions)
Jump to: navigation, search
Line 82: Line 82:
# Glasser and Van Essen (J. Neuroscience, 2011)
# Glasser and Van Essen (J. Neuroscience, 2011)
#Yeo et al. (J. Neurophysiol., 2011)
#Yeo et al. (J. Neurophysiol., 2011)
#[http://www.humanconnectome.org/ ]  
#[http://www.humanconnectome.org/ '''http://www.humanconnectome.org''']  
# Van Essen et al., (NeuroImage, 2012)
# Van Essen et al., (NeuroImage, 2012)
#Van Essen (NeuroImage, 2005)
#Van Essen (NeuroImage, 2005)

Revision as of 18:56, 21 March 2012

Conte69 Atlas

An improved human surface-based atlas and associated reference data are now available. The ‘Conte69’ atlas was generated from 69 healthy adults registered to the “164k_fs_LR” surface mesh (Refs 1, 2).

The atlas dataset includes Population-average data derived from the Conte69 subjects:

  • Average midthickness surfaces, in register between left and right hemispheres (Ref 1).
  • Average cortical myelin maps (Ref 2).
  • Average cortical thickness, with and without curvature correction.
  • Average T1w and T2w volumes, registered using FSL’s nonlinear FNIRT.

Reference data mapped to the Conte69 atlas:

  • A composite cortical parcellation based on 52 surface-mapped areas (Ref 1).
  • The classical Brodmann parcellation (Ref 1)
  • Cortical networks derived from a resting-state functional connectivity analysis (Ref 3).

More to come. Standard Freesurfer-processed datasets can be mapped to the fs_LR mesh and viewed on the Conte69 atlas using a ‘freesurfer_to_fs_LR’ script to be released soon (see below). Also, data from the Human Connectome Project (Refs 4, 5) will be registered to the fs_LR mesh as part of the standard processing pipeline.

How do I access the data? The data are available in SumsDB (via

[1]). Read the next section to decide which particular archives to download.

Choice of surface mesh. Two versions of the data are available that differ in the resolution of the surface mesh (Freesurfer vs PALS).

The Conte69_164k_fs_LR atlas has higher spatial resolution and is a step closer to the primary Freesurfer data [1]. [‘164k’ is shorthand for 163,842 surface nodes).]

The Conte69_74k_pals atlas enables immediate comparison with existing data mapped to the PALS-B12 atlas [6]. [‘74k’ is shorthand for 73,730 surface nodes).]

Conversion of existing data from the 164k_fs_LR mesh to the 74k_pals mesh and vice versa can be readily done by a single command-line step (see below).

  • Download and view the Conte69_164k_fs_LR atlas. Link to [2] then click on the ‘disk’ icon for atlas_Conte69_164k_fs_LR.zip.
    • Unpack the archive.
    • Change to the atlas_Conte69_164k_fs_LR/ directory.
    • Launch caret5 and select the spec file for the left and/or right hemisphere (Conte69_atlas.L.164k_fs_LR.c5.spec, Conte69_atlas.R.164k_fs_LR.c5.spec)
    • Press Load Scenes, then view the various scenes (highlight and double-click).
    • For additional datasets change to the Extra_Conte69_164k_fs_LR/ directory and open one or both ‘long’ spec files that include additional files and scenes (Conte69_atlas.L.long.164k_fs_LR.c5.spec, Conte69_atlas.R.long.164k_fs_LR.c5.spec
  • 74k_pals mesh. Data from the 164k_fs_LR mesh have been registered to the PALS-B12 atlas, which uses a ‘74k_pals’ mesh (shorthand for 73,730 nodes, [6]). To download this, link to [3] then click on the ‘disk’ icon for *atlas_Conte69_74k_pals.zip*
    • Unpack the archive.
    • Change to the atlas_Conte69_74k_fs_LR/ directory.
    • Launch caret5, open the for the left and/or right hemisphere spec file (Conte69_atlas.L.74k_fs_LR.c5.spec, Conte69_atlas.L.74k_fs_LR.c5.spec)
    • Press Load Scenes, then view the various scenes (highlight and double-click).

How can I convert (register) additional data between 164k_fs_LR and 74k_pals meshes?

  • Download and unzip inter-atlas_deformation_maps.zip archive: [4]
  • In a terminal window, change (cd) to whichever target directory (fs_L, fs_LR, fs_R, pals_L, pals_R) represents the mesh you desire for the *output* files.
  • Enter “caret_command –deformation-map-apply”, then follow the displayed command-line instructions in order to register the files of interest.

How can I convert (register) standard Freesurfer data to the 164k_fs_LR mesh? Data that have been processed by standard Freesurfer ‘recon-all’ (v5.1 only? Or 5.0 and 5.1?) can be converted to the fs_LR mesh using a ‘freesurfer_to_fs_LR’ script and associated dataset available at [5] - Download Freesurfer_to_fs_LR Pipeline Distribution. Instructions are at: [6]. This script will let you convert existing or future Freesurfer datasets to the fs_LR format. This facilitates interhemispheric comparisons in your own data (e.g., symmetries and asymmetries) as well as comparisons with the Conte69 atlas datasets provided herein.

Additional notes. Improved atlas surfaces and myelin maps. The Conte69 atlas dataset was derived from Freesurfer segmentation, surface reconstruction, and registration to the fsaverage left and right hemispheres, followed by interhemispheric registration to the left-right hybrid atlas, bringing the two hemispheres into geographic correspondence (Ref 1). The current data release benefits from several refinements in surface generation and in the myelin mapping algorithm compared to the published data (Refs 1, 2).

Connectome Workbench data format. The Conte69 atlas dataset is also available in the new Connectome Workbench data format being developed for the Human Connectome Project. A beta-release of Connectome Workbench will be made available in the summer of 2012.


  1. Van Essen et al., (Cerebral Cortex, 2011)
  2. Glasser and Van Essen (J. Neuroscience, 2011)
  3. Yeo et al. (J. Neurophysiol., 2011)
  4. http://www.humanconnectome.org
  5. Van Essen et al., (NeuroImage, 2012)
  6. Van Essen (NeuroImage, 2005)
Personal tools
Sums Database