Neuroimaging Pattern Masks
June 2, 2026 · View on GitHub
This docs/ directory is the documentation hub for the
canlab/Neuroimaging_Pattern_Masks
repository. It indexes every study folder, links to per-folder documentation
(contents_description.md), and explains how to load each map with the
CANlab Core Tools.
GitHub renders this page automatically when you open the
docs/folder. The file is also compatible with GitHub Pages (Jekyll) — to enable a public site, turn on Pages with sourcedocs/in repository settings.
For the project-level overview, see the top-level README.
What is in this repository
The repository collects three kinds of brain maps, each with its own top-level landing page:
| Type | Top-level folder | Landing page |
|---|---|---|
| Multivariate signatures — pre-trained predictive patterns | Multivariate_signature_patterns/ | README |
| Atlases & parcellations — labeled regions / networks | Atlases_and_parcellations/ | README |
| CANlab meta-analyses — consensus maps across studies | CANlab_Meta_analysis_maps/ | README |
| Neurosynth maps | Neurosynth_maps/ | README |
| Individual-study maps | Individual_study_maps/ | README |
| Spatial basis functions — continuous gradients / ICA components | spatial_basis_functions/ | README |
| Templates — reference T1 templates | templates/ | README |
A web companion with figures and references is hosted at the CANlab Brain Patterns site.
Loading maps
Most maps can be loaded by keyword through CanlabCore:
% Multivariate signatures (returns fmri_data + names)
[obj, networknames, imagenames] = load_image_set('siips'); % SIIPS1
[obj, networknames, imagenames] = load_image_set('npsplus'); % NPS, SIIPS, PINES, Rejection, VPS, GSR...
[obj, networknames, imagenames] = load_image_set('painsig'); % NPS + SIIPS only
% Atlases (returns atlas object)
atl = load_atlas('canlab2024');
atl = load_atlas('canlab2023');
atl = load_atlas('glasser_fmriprep20');
atl = load_atlas('schaefer400');
See load_image_set.m and load_atlas.m in CanlabCore for the full keyword list.
Applying signatures to new data
The repo ships two convenience wrappers at the root of
Multivariate_signature_patterns/:
-
apply_all_signatures.m— applies a configurable set of signatures (NPS, SIIPS, PINES, VPS, Rejection, GSR, HR, …) to a cell array offmri_dataobjects and returns a results table. Supportssimilarity_metric(dot-product / cosine / correlation),image_scaling(none / center / z-score / l2-norm), and a namedimage_setto choose which signature bundle to apply. Example:[SIG, sigtable] = apply_all_signatures(DATA_OBJ, ... 'similarity_metric', 'cosine_similarity', ... 'image_scaling', 'l2norm_images', ... 'image_set', 'npsplus'); -
apply_siips.m— SIIPS1-specific wrapper that accepts wildcards, filename lists, orfmri_dataobjects and returns whole-signature responses plus local responses for each of the 44 FDR-thresholded SIIPS subregions.
Both helpers internally call CanlabCore's apply_mask /
canlab_pattern_similarity, but they are easier to drop into a batch
pipeline than wiring those up by hand.
Per-study documentation
Each study folder contains:
- A primary-reference PDF named
<lastauthor>_<year>_<keywords>.pdf(where available). - A
contents_description.mdthat describes the folder (overview, key images, loading instructions, file inventory, citations). - A
visualize_contents.mMATLAB script that regenerates cortical-surface, isosurface (where relevant), and montage figures intopng_images/. - The image files themselves (NIfTI /
atlas.matobjects / signature.matobjects). - Any existing author-authored
README.mdis preserved verbatim and linked fromcontents_description.md.
Shared rendering helpers used by every per-folder script live in
docs/:
canlab_render_patterns.m— surface + montage + isosurface forfmri_datapatterns.canlab_render_atlas.m— montage + isosurface foratlasobjects.run_all_multivariate.m— batch-re-renders every folder in a chosen category.
Multivariate signatures — all entries documented
Atlases and parcellations — all entries documented
| Year | Folder | Description |
|---|---|---|
| 2006 | Desikan-Killiany | Desikan-Killiany cortical atlas |
| 2009 | Destrieux | Destrieux gyral/sulcal cortical atlas |
| 2011 | Buckner 7networks | 7 resting-state networks |
| 2011 | Yeo 17networks | 17 resting-state networks |
| 2012 | Desikan-Killiany-Tourville (DKT) | DKT cortical atlas |
| 2013 | Shen/Constable 268 | 268-node functional parcellation |
| 2014 | Keuken 7T subcortex | 7T high-resolution subcortex |
| 2016 | CIT168 amygdala v1.0.3 | Amygdala subnuclei |
| 2016 | Fan Brainnetome 273 | Brainnetome parcellation |
| 2016 | Glasser HCP-MMP1 | 360 multimodal cortical parcels (pilot) |
| 2018 | CIT168 reinf-learn v1.0.0 | CIT168 subcortical v1.0.0 |
| 2018 | CIT168 reinf-learn v1.1.0 | CIT168 subcortical v1.1.0 |
| 2018 | Iglesias thalamic | Iglesias thalamic nuclei |
| 2018 | Schaefer/Yeo multires | 100–1000 parcel multi-resolution cortex |
| 2018 | Wager combined atlas | Original CANlab whole-brain composite |
| 2019 | Cartmell NAc core/shell | NAcc core / shell |
| 2019 | Kragel PAG | Periaqueductal gray subregions |
| 2019 | Wager pain pathways | Canonical ascending/descending pain |
| 2019 | Zheng pain connectivity modules | Pain connectivity modules |
| 2020 | Iglesias hypothalamus | Hypothalamic subregions |
| 2020 | JulichBrain v3.0.3 | Cytoarchitectonic |
| 2020 | Pandora WM atlas | White-matter bundles |
| 2020 | Thiebaut de Schotten WM | White-matter atlas |
| 2020 | Tian subcortical v1.1 | Multi-scale subcortex |
| 2022 | Hansen PET tracer maps | Neurotransmitter PET maps |
| 2023 | Bianciardi Brainstem Navigator v0.9 | Brainstem nuclei |
| 2023 | CANlab atlas (2023) | CANlab combined atlas (2023) |
| 2023 | CANlab atlas CIFTI | CIFTI grayordinate version |
| 2023 | Harvard AAN brainstem | Ascending arousal network |
| 2023 | Levinson/Bari limbic brainstem | Limbic brainstem |
| 2024 | CANlab atlas (2024) | Latest CANlab combined atlas |
| 2026 | Pourmajidian mitochondrial energetic capacity | Mitochondrial energetic capacity |
| util | readme_and_summary | Cross-atlas summary materials |
CANlab meta-analysis maps — all entries documented
Neurosynth maps
- 2016 Neurosynth 100 topics (v4/v5) — pilot
- mkda/ — MKDA per-term kernel-density maps
- scripts/ — Helper scripts
- topic_terms_csv/ — Per-topic term-frequency CSVs
Individual-study maps
- 2024 Bo — Emotion regulation Bayes factor (Nat Neurosci) — pilot
- 2026 Miao — Social/ToM Bayes factor
Spatial basis functions
- Margulies — Principal cortical gradient (PNAS 2016) — pilot
- HCP 91k
- HCP group ICAs
- Transcriptomic gradients
- Mitochondrial profile maps (Mosharov 2025)
Templates
- templates landing page — overview of MNI templates included.
- cerebellum/
- transforms/
Conventions used in this documentation
Per the contribution guidelines, every study folder
follows the same layout. The contents_description.md in each folder is the
canonical entry point and contains:
- Overview — what the resource is, in a few lines, with the primary reference linked at the publisher's site (and citations to any secondary references found in the folder).
- Key images — 1–2 representative figures (montage, surface, or isosurface) taken
from the CANlab Brain Patterns site or generated by
visualize_contents.m. Pre-rendered figures already present in folders (e.g.,images_and_regions/,figures/,png_images/) are embedded directly. - How to load — the exact
load_atlas()/load_image_set()call, or the filename to pass tofmri_data()/atlas()directly. - Construction scripts — links to constructor scripts in the folder or in CanlabCore.
- File inventory — every NIfTI /
.mat/.min the folder with one line each. Individual.pngfiles are not listed (they are embedded as key images instead). - Citations — primary and secondary papers, plus links.
Existing author-authored README.md / Readme.md / Readme.txt files in
sub-folders are preserved verbatim and linked from the new
contents_description.md, mirroring the Glasser-folder pattern.
Regenerating figures
After modifying any folder, regenerate its PNGs:
cd /path/to/Atlases_and_parcellations/2016_Glasser_Nature_HumanConnectomeParcellation
visualize_contents(); % writes png_images/<label>_{surface,montage,isosurface}.png
To re-render every signature in one MATLAB session:
addpath('/path/to/Neuroimaging_Pattern_Masks/docs');
run_all_multivariate('Multivariate_signature_patterns'); % or any other category
Adding a new study folder
- Create the folder under the appropriate category with name
YYYY_<lastauthor>_<keywords>. - Drop in the NIfTI /
.matfiles plus a copy of the primary reference PDF (filename:<lastauthor>_<year>_<keywords>.pdf). - Copy a
visualize_contents.mfrom a similar folder and adapt theimgscell. - Run it once locally to generate
png_images/. - Write
contents_description.mdfollowing the sections above. - Add the entry to the appropriate category list in the top-level
landing page (e.g.,
Atlases_and_parcellations/README.md) and to the master list in thisdocs/README.md. - If applicable, add a keyword to the switch in
CanlabCore/Data_extraction/load_image_set.morload_atlas.m.
Documentation status
Every study folder in the repository now has a contents_description.md
and (in nearly all cases) a visualize_contents.m. Top-level landing
pages list each subfolder and link to its docs.
PDF coverage: 30+ PDFs are checked in directly. For a few paywalled papers (Reddan 2018 Neuron, Speer 2023 J Neurosci, Coll 2022, Kragel 2015 SCAN) the DOI link is recorded but no local PDF is present.
MATLAB / PNG coverage: every visualize_contents.m writes its
output to png_images/. The full batch is reproducible by running
docs/run_all_multivariate.m once per
category.
If a folder you need is missing something, the underlying image files
and any pre-existing README.md / create_*.m script in that folder
remain authoritative — the new docs are an additional indexing /
discovery layer, not a replacement.