Atlases, named regions, and signature patterns

May 7, 2026 · View on GitHub

CanlabCore consumes a curated collection of brain atlases / parcellations, named regions of interest, multivariate signature patterns, and meta-analytic maps. The image files themselves are distributed via the canlab/Neuroimaging_Pattern_Masks repository, while the keyword registries that resolve a short name to a specific saved object live inside CanlabCore. Three loader functions are the main entry points:

  • load_atlas('keyword') returns an atlas object (a labeled parcellation with .dat, .probability_maps, .labels, .label_descriptions, and .references). See atlas_methods.md for what you can do with it.
  • load_image_set('keyword') returns an fmri_data object holding one or more multivariate signature patterns or meta-analytic maps, ready to apply with apply_mask, canlab_pattern_similarity, or image_similarity_plot. The full keyword table for load_image_set (datasets and signatures) lives in sample_datasets.md.
  • canlab_load_ROI('region_name') returns a region object for a single named region (e.g. 'pag', 'lc', 'nacc'), drawn from hand-picked sources across atlases and individual papers. Optionally also returns a 1 mm atlas object covering that region.

For curated descriptions, figures, and usage notes, see the companion CANlab Brain Patterns site. The main toolbox documentation is at canlab.github.io.

The atlas keywords typically accept suffixes that select the reference space (_fmriprep20 for MNI152NLin2009cAsym, _fsl6 for MNI152NLin6Asym) and sampling resolution (_1mm, _2mm); some accept granularity (_fine, _coarse). Defaults vary by atlas — see the docstring of load_atlas for the full list.

Atlases and parcellations

Whole-brain combined atlases

The canlab2018, canlab2023, canlab2024, and opencanlab2024 atlases are dynamically assembled whole-brain parcellations that combine several published atlases into a single, internally consistent labeling. Cortex is taken from the Glasser HCP-MMP1 parcellation, subcortex from CIT168 (and Pauli striatum / Iglesias thalamus in newer builds), thalamus from Iglesias and/or Morel, cerebellum from SUIT (Diedrichsen), and brainstem nuclei from Bianciardi (canlab2023/2024, which therefore require local assembly because the Bianciardi atlas cannot be redistributed). opencanlab2024 is a fully open-license variant that omits Bianciardi (synthesizing approximations of some brainstem nuclei) and ships pre-assembled. canlab2018 is the legacy build, deprecated in favor of canlab2023. See the README of the Neuroimaging_Pattern_Masks repository for the per-region provenance table.

Cortical parcellations

The Glasser HCP-MMP1 multimodal parcellation defines 180 areas per hemisphere from cortical myelin, thickness, function, and connectivity in the HCP cohort. CanlabCore ships volumetric projections built with registration fusion to the fmriprep (MNI152NLin2009cAsym) and FSL (MNI152NLin6Asym) reference spaces. Glasser, M.F. et al. (2016). A multi-modal parcellation of human cerebral cortex. Nature, 536, 171-178.

The Desikan-Killiany atlas labels 34 cortical gyri per hemisphere based on sulcal anatomy and is the default Freesurfer cortical labeling. Desikan, R.S. et al. (2006). An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. NeuroImage, 31(3), 968-980.

The DKT (Desikan-Killiany-Tourville) atlas refines the Desikan-Killiany boundaries to be more anatomically consistent across labs. Klein, A. & Tourville, J. (2012). 101 labeled brain images and a consistent human cortical labeling protocol. Frontiers in Neuroscience, 6, 171.

The Destrieux atlas (also distributed with Freesurfer) divides the cortex into ~74 gyral and sulcal parcels per hemisphere, defined by local curvature. Destrieux, C., Fischl, B., Dale, A. & Halgren, E. (2010). Automatic parcellation of human cortical gyri and sulci using standard anatomical nomenclature. NeuroImage, 53(1), 1-15.

The Schaefer 2018 / Yeo 17 networks parcellation provides 100-1000 cortical parcels (CanlabCore ships the 400-parcel and 17-network versions) optimized for local gradient and global similarity in resting-state fMRI. Schaefer, A. et al. (2018). Local-global parcellation of the human cerebral cortex from intrinsic functional connectivity MRI. Cerebral Cortex, 28(9), 3095-3114. The 17-network labeling is from Yeo, B.T.T. et al. (2011). The organization of the human cerebral cortex estimated by intrinsic functional connectivity. Journal of Neurophysiology, 106(3), 1125-1165.

The Shen 268 parcellation is a functional connectivity-based atlas covering cortex, subcortex, and cerebellum (CanlabCore ships fmriprep and FSL projections in addition to the original Colin27 space). Shen, X., Tokoglu, F., Papademetris, X. & Constable, R.T. (2013). Groupwise whole-brain parcellation from resting-state fMRI data for network node identification. NeuroImage, 82, 403-415.

Subcortical atlases

The CIT168 subcortical atlas is a probabilistic parcellation of 16 subcortical structures (basal ganglia, amygdala, thalamic and diencephalic nuclei) derived from high-resolution multi-contrast 7T templates. CanlabCore distributes v1.1.0 in fmriprep and FSL spaces, plus a separate amygdala-only sub-atlas (v1.0.3). Pauli, W.M., Nili, A.N. & Tyszka, J.M. (2018). A high-resolution probabilistic in vivo atlas of human subcortical brain nuclei. Scientific Data, 5, 180063. The CIT168 amygdala parcellation is described in Tyszka, J.M. & Pauli, W.M. (2016). In vivo delineation of subdivisions of the human amygdaloid complex in a high-resolution group template. Human Brain Mapping, 37(11), 3979-3998.

The Pauli striatum atlas decomposes the striatum into five probabilistic functional zones using diffusion-derived connectivity. Pauli, W.M., O'Reilly, R.C., Yarkoni, T. & Wager, T.D. (2016). Regional specialization within the human striatum for diverse psychological functions. PNAS, 113(7), 1907-1912.

The Brainnetome atlas parcellates cortex and subcortex into 246 regions defined by anatomical connectivity from diffusion MRI. Fan, L. et al. (2016). The Human Brainnetome Atlas: A new brain atlas based on connectional architecture. Cerebral Cortex, 26(8), 3508-3526.

Thalamus

The Morel / Krauth thalamic atlas is a histology-based stereotaxic parcellation of human thalamic nuclei. Krauth, A. et al. (2010). A mean three-dimensional atlas of the human thalamus: generation from multiple histological data. NeuroImage, 49(3), 2053-2062. The original histological work is Morel, A., Magnin, M. & Jeanmonod, D. (1997). Multiarchitectonic and stereotactic atlas of the human thalamus. Journal of Comparative Neurology, 387(4), 588-630.

The Iglesias thalamus atlas is a probabilistic Bayesian segmentation of 26 thalamic nuclei using ex vivo histology and in vivo MRI; the CanlabCore build is projected to fmriprep and FSL spaces. Iglesias, J.E. et al. (2018). A probabilistic atlas of the human thalamic nuclei combining ex vivo MRI and histology. NeuroImage, 183, 314-326.

The Tian 3T subcortex atlas provides four levels of granularity (S1-S4, 16-54 regions) of subcortical parcellation from 3T resting-state fMRI gradients, and is available in both fmriprep and FSL spaces. Tian, Y., Margulies, D.S., Breakspear, M. & Zalesky, A. (2020). Topographic organization of the human subcortex unveiled with functional connectivity gradients. Nature Neuroscience, 23(11), 1421-1432.

Hypothalamus

The Iglesias / Billot hypothalamus segmentation uses a CNN to delineate five hypothalamic subregions from T1 MRI. Billot, B. et al. (2020). Automated segmentation of the hypothalamus and associated subunits in brain MRI. NeuroImage, 223, 117287.

Brainstem and cerebellum

The SUIT cerebellar atlas (Diedrichsen) is a high-resolution lobular parcellation of the cerebellum aligned to a dedicated cerebellar template. Diedrichsen, J., Balsters, J.H., Flavell, J., Cussans, E. & Ramnani, N. (2009). A probabilistic MR atlas of the human cerebellum. NeuroImage, 46(1), 39-46. The functional/lobular parcellation used in newer SUIT releases is Diedrichsen, J. & Zotow, E. (2015). Surface-based display of volume-averaged cerebellar imaging data. PLoS ONE, 10(7), e0133402.

The Bianciardi mesopontine atlas provides probabilistic templates for ~30 brainstem nuclei built from 7T multimodal in vivo data. Distribution is restricted, so CanlabCore assembles a derivative atlas locally on first use. Bianciardi, M. et al. (2018). A probabilistic template of human mesopontine tegmental nuclei from in vivo 7T MRI. NeuroImage, 170, 222-230. See also the project page at the Brainstem Imaging Lab.

The Harvard Ascending Arousal Network (AAN) v2.0 atlas labels brainstem and forebrain nuclei central to arousal and consciousness, based on histology, immunohistochemistry, and diffusion tractography. Edlow, B.L. et al. (2012). Neuroanatomic connectivity of the human ascending arousal system critical to consciousness and its disorders. Journal of Neuropathology and Experimental Neurology, 71(6), 531-546. v2.0 is distributed via the Center for Neurotechnology and Neurorecovery AAN page.

The Levinson-Bari Limbic Brainstem Atlas is a probabilistic, openly licensed atlas of brainstem limbic structures including VTA, dorsal raphe, locus coeruleus, NTS, and PAG. Levinson, S., Miller, M., Iftekhar, A., Justo, M., Arriola, D., Wei, W., Hazany, S., Avecillas-Chasin, J.M., Kuhn, T.P., Horn, A. et al. (2023). A structural connectivity atlas of limbic brainstem nuclei. Frontiers in Neuroimaging, 1, 1009399.

The Kragel 2019 PAG atlas (kragel2019pag keyword) is a 7T-derived sub-parcellation of the periaqueductal gray into dorsal, lateral, and ventrolateral columns via unsupervised k-means clustering of voxel positions. Kragel, P.A., Bianciardi, M., Hartley, L., Matthewson, G., Choi, J.-K., Quigley, K.S., Wald, L.L., Wager, T.D., Barrett, L.F., & Satpute, A.B. (2019). Functional involvement of human periaqueductal gray and other midbrain nuclei in cognitive control. Journal of Neuroscience, 39(31), 6180-6189.

7T high-resolution

The Keuken 7T atlas is a probabilistic atlas of subcortical structures (STN, SN, RN, GPe, GPi, thalamus) built from 7T quantitative MRI in 30 healthy adults. Keuken, M.C. et al. (2014). Quantifying inter-individual anatomical variability in the subcortex using 7T structural MRI. NeuroImage, 94, 40-46.

Functional networks

The Buckner / Yeo 2011 networks atlas provides 7- and 17-network cortical parcellations from resting-state fMRI in 1000 subjects. CanlabCore ships an atlas object that combines the cortical networks with associated cerebellar (Buckner et al. 2011) and striatal (Choi et al. 2012) functional networks. Yeo, B.T.T. et al. (2011). The organization of the human cerebral cortex estimated by intrinsic functional connectivity. Journal of Neurophysiology, 106(3), 1125-1165. Buckner, R.L. et al. (2011). The organization of the human cerebellum estimated by intrinsic functional connectivity. Journal of Neurophysiology, 106(5), 2322-2345. Choi, E.Y., Yeo, B.T.T. & Buckner, R.L. (2012). The organization of the human striatum estimated by intrinsic functional connectivity. Journal of Neurophysiology, 108(8), 2242-2263.

Pain pathways

The CANlab pain pathways atlas is a curated combination of structures involved in nociceptive processing — spinothalamic and trigeminothalamic targets in thalamus, somatosensory cortex, posterior and middle insula, dorsal posterior cingulate, amygdala, hypothalamus, PAG, parabrachial complex, and rostral ventromedial medulla. It is documented and figured on the CANlab Brain Patterns: pain pathways page. The companion paper for the underlying signature work is Wager, T.D. et al. (2013). An fMRI-based neurologic signature of physical pain. New England Journal of Medicine, 368, 1388-1397. A 2024 refresh (painpathways2024) re-bases the pathway on canlab2024 parcels and adds finer brainstem coverage; see the Neuroimaging_Pattern_Masks README.

Other atlases

The Faillenot insular atlas divides the human insula into three cytoarchitectonic gyri. Faillenot, I., Heckemann, R.A., Frot, M. & Hammers, A. (2017). Macroanatomy and 3D probabilistic atlas of the human insula. NeuroImage, 150, 88-98.

The Cartmell NAc core/shell atlas provides probabilistic delineations of nucleus accumbens core and shell. Cartmell, S.C.D. et al. (2019). Multimodal characterization of the human nucleus accumbens. NeuroImage, 198, 137-149.

The de la Vega Neurosynth atlas is a meta-analytically derived co-activation parcellation of medial frontal cortex (and a related whole-cortex variant) from automated meta-analysis on Neurosynth. de la Vega, A., Chang, L.J., Banich, M.T., Wager, T.D. & Yarkoni, T. (2016). Large-scale meta-analysis of human medial frontal cortex reveals tripartite functional organization. Journal of Neuroscience, 36(24), 6553-6562. The companion Neurosynth platform is documented at neurosynth.org.

The Julich-Brain cytoarchitectonic atlas is a probabilistic histology-based parcellation covering most of cortex and many subcortical structures. Amunts, K., Mohlberg, H., Bludau, S. & Zilles, K. (2020). Julich-Brain: A 3D probabilistic atlas of the human brain's cytoarchitecture. Science, 369(6506), 988-992. Live updates and download links are at the Julich-Brain page on EBRAINS.

Named regions (canlab_load_ROI)

Named regions are hand-picked single-region masks selected from atlases or individual papers, returned as region objects suitable for display (addbrain, montage) or extraction (extract_roi_averages). Regions are grouped here by anatomy.

Cortex

KeywordDescriptionSource
vmpfcVentromedial prefrontal + posterior cingulate, midlineHand-drawn (Tor Wager)

Forebrain (non-basal ganglia)

KeywordDescriptionSource
nacc, nacNucleus accumbensPauli et al. 2018 (CIT168)
amygdalaAmygdalaTyszka & Pauli 2016 (CIT168)
hippHippocampusEickhoff et al. 2005 (SPM Anatomy Toolbox via canlab2018)
BST, BNST, SLEABed nucleus of stria terminalis / sublenticular extended amygdalaPauli et al. 2018

Basal ganglia

KeywordDescriptionSource
caudate, cauCaudate nucleusPauli et al. 2018 (CIT168)
put, putamenPutamenPauli et al. 2018 (CIT168)
gpGlobus pallidus (GPe + GPi)Keuken et al. 2014
GPeGlobus pallidus externaKeuken et al. 2014
GPiGlobus pallidus internaKeuken et al. 2014
VeP, vep, vpallVentral pallidumPauli et al. 2018

Thalamus, diencephalon, epithalamus

KeywordDescriptionSource
thalamus, thalThalamus main bodyKrauth et al. 2010 (Morel)
mdMediodorsal thalamusKrauth et al. 2010
cmCentromedian thalamusKrauth et al. 2010
lgnLateral geniculate nucleusKrauth et al. 2010
mgnMedial geniculate nucleusKrauth et al. 2010
VPthal, VPLVentral posterior thalamusKrauth et al. 2010
intralaminar_thalIntralaminar / midline thalamusKrauth et al. 2010
stnSubthalamic nucleusKeuken et al. 2014
habenula, HNHabenulaPauli et al. 2018
mammillary, mammMammillary bodiesPauli et al. 2018
hypothalamus, hy, hythalHypothalamusPauli et al. 2018 (CIT168)

General brainstem

KeywordDescriptionSource
brainstemWhole brainstem (cleaned SPM8 tissue probability map)Hand-cleaned (Tor Wager), SPM8 TPM
midbrainWhole midbrainCarmack et al. 2004

Midbrain

KeywordDescriptionSource
pagPeriaqueductal gray (hand-drawn, aqueduct masked)Tor Wager 2018; aqueduct from Keuken et al. 2014
scSuperior colliculus (hand-drawn)Tor Wager 2018; aqueduct from Keuken et al. 2014
icInferior colliculus (hand-drawn)Tor Wager 2018; aqueduct from Keuken et al. 2014
drnDorsal raphe nucleusBeliveau et al. 2015
mrnMedian raphe nucleusBeliveau et al. 2015
ncfNucleus cuneiformisZambreanu et al. 2005
PBPParabrachial pigmented nucleusPauli et al. 2018
snSubstantia nigraKeuken et al. 2014
SNc, sncSubstantia nigra compactaPauli et al. 2018
SNr, snrSubstantia nigra reticularisPauli et al. 2018
VTA, vtaVentral tegmental areaPauli et al. 2018
rnRed nucleusKeuken et al. 2014

Pons

KeywordDescriptionSource
pbnParabrachial complex (current, from canlab2024)Bianciardi et al. 2018 via canlab2024
pbn_oldParabrachial complex (legacy)Fairhurst et al. 2007
lc, locus_coeruleusLocus coeruleus (2-SD probability mask)Keren et al. 2009
nrp_B5Nucleus raphe pontis (B5)Bär et al. 2016

Medulla

KeywordDescriptionSource
rvm, rvm_brooksRostral ventromedial medullaBrooks et al. 2017
rvm_oldRostral ventromedial medulla (legacy hand-drawn)Tor Wager (legacy)
nrm, raphe magnusNucleus raphe magnusBär et al. 2016
medullary_rapheMedullary rapheNash et al. 2009
ncs_B6_B8Nucleus centralis superior (B6/B8)Bär et al. 2016
nuc_ambiguusNucleus ambiguusSclocco et al. 2016
dmnx_nts, ntsDorsal motor nucleus of vagus / nucleus of the solitary tractSclocco et al. 2016
spinal_trigeminalSpinal trigeminal nucleusNash et al. 2009
oliveInferior olive (binary mask)Hand-drawn / legacy

Multivariate signature patterns

CanlabCore distributes a curated set of multivariate brain patterns — spatially-distributed weight maps that, when applied to a new image (typically with apply_mask plus a dot product, or with canlab_pattern_similarity / apply_nps / apply_all_signatures), yield a scalar prediction for a psychological or clinical state. The full keyword table for load_image_set (signatures plus sample fMRI datasets) is in sample_datasets.md. A few of the most prominent signatures and their original publications:

Many other signatures (e.g. for autonomic arousal, reward, empathy, threat, working memory, response conflict) are registered in load_image_set. See sample_datasets.md for the exhaustive table.

References to other sources