Cortical Atlas Parcellations in MNI space

Below you find information about how and where to obtain cortical atlas parcellations suitable for fMRI/dMRI based connectomic analyses using Lead-DBS.   If an atlas you know of is missing, please contact us. Also, we are interested in distributing cortical parcellation atlases preinstalled within Lead-DBS, if possible.

Looking for subcortical atlases in MNI space? Please see this page.

Mindboggle 101 (Klein 2012)

The Mindboggle 101 dataset consists of 101 labeled brain images that have been manually labeled largely following the Desikan protocol. It also consists of a group-level parcellation atlas which has been included into Lead-DBS for connectomic analyses.

Cortical Area Parcellation from Resting-State Correlations (Gordon 2014)

The Cortical Area Parcellation from Resting-State Correlations dataset consists of 333 cortical patches segmented using resting-state fMRI (Gordon 2014).

MICCAI 2012 Multi-Atlas Labeling Workshop and Challenge (Neuromorphometrics)

The MICCAI 2012 Multi-Atlas Labeling Workshop and Challenge dataset is a manually segmented whole-brain parcellation (see refs).

How to obtain the atlas:

Brainnetome Atlas parcellation (Fan 2016)

The Brainnetome atlas is an in vivo map based on fMRI and dMRI, with more fine-grained functional brain subregions and detailed anatomical and functional connection patterns for each area. Currently, the Brainnetome atlas contains 246 subregions of the bilateral hemispheres.

Automated Anatomical Labeling (Tzourio-Mazoyer 2002)

The AAL atlas is probably the most widely used cortical parcellation map in connectomic literature.
On 7th August 2015, an updated version (AAL2) has been released.

Harvard-Oxford cortical/subcortical atlases (Makris 2006)

A frequently applied cortical atlas motivated by macroanatomical boundaries.

AICHA: An atlas of intrinsic connectivity of homotopic areas (Joliot 2015)

The AICHA atlas – a detailed cortical atlas based on functional connectivity in 281 subjects.

Hammersmith atlas (Hammers 2003, Gousias 2008)

Adult brain maximum probability map (“Hammersmith atlas”; n30r83) in MNI space.

Yeo 2011 functional parcellations (Yeo 2011)

fMRI atlas based on 1000 subjects exhibiting co-activations of the brain. Two versions are available that include 7 vs. 17 neworks.

HCP MMP 1.0 (Glasser 2016)

Probably the most detailed cortical in-vivo parcellation yet, the HCP MMP 1.0 has been built using surface-based registrations of multimodal MR acquisitions from 210 HCP subjects. 180 areas per hemisphere have been identified. Please note that due to the nature of how this parcellation has been built, it may not be ideally suited for usage within Lead-DBS.

JuBrain / Juelich histological atlas (Eickhoff 2005)

A probabilistic atlas created by averaging multi-subject post-mortem cyto- and myelo-architectonic segmentations, performed by the team of Profs Zilles and Amunts in Jülich and Düsseldorf. Cytoarchitectonic areas were analyzed in histological sections of ten human postmortem brains. The maps are based on image analysis and statistical criteria for localizing areal borders. Cytoarchitectonic maps have been developed during the past 20 years as a joint effort of many doctoral students, post docs and guest scientists. Maps, which have been published, are available for the scientific community.

PrAGMATiC (Huth 2016)

Advanced and modern functional atlas based on task fMRI.

Desikan-Killiany Atlas (Desikan 2006)

A parcellation scheme widely used in the freesurfer world subdividing the human cerebral cortex on MRI scans into gyral based regions of interest.

Destrieux Atlas (Destrieux 2010)

A parcellation scheme widely used in the freesurfer world based on sulcal depth and yielding precise automated definition of cortical gyri and sulci.

MarsAtlas (Auzias 2016)

A cortical parcellation model based on macroanatomical information.

fMRI-based random parcellations (Craddock 2011)

Fine-grained random parcellations informed by rs-fMRI data.

Voxelwise parcellations (Lead-DBS)

Three voxel-wise parcellations are supplied with Lead-DBS and have been built to create standardized connectivity matrices that exhibit off-diagonal elements that appear as parallel lines to the main diagonal (inter-hemispheric connections between homologous regions). Three versions (35 thousand, 15 thousand and 8 thousand nodes) are pre-installed within Lead-DBS. Together with advanced normalization algorithms (such as multimodal ANTs defaults and DARTEL pipelines), these connectivity matrices allow for high-definition connectomic analyses in parcellation schemes that are comparable across studies.

How to obtain the atlas:

  • These parcellation schemes come preinstalled within Lead-DBS.

Related citations:

  • Parcellation schemes are not described in a separate publication. Please cite Lead-DBS software including version and webpage to describe the data used in a reproducible fashion.