neurodata / neuroparc

Neuroparc

• Atlas Spec
• Table

This repository contains a number of useful parcellations, templates, masks, and transforms to (and from) MNI152NLin6 space. The files are named according to the BIDs specification

Atlas Info Summary

Atlas Name	# Regions	Symmetrical?	Hierarchical?	Labelled?	Generation Method	Average Vol/Region	Native coordinate space	Description	Reference Publication	Year of Origin	File Source/Download URL
AAL	116	No	No	Yes	Delineated with respect to anatomical landmarks (following sulci course in brain)	12758.353	MNI	Automated anatomical labelling based on sulci.	https://www.ncbi.nlm.nih.gov/pubmed/11771995	2002	https://www.gin.cnrs.fr/en/tools/aal/ http://www.gin.cnrs.fr/wp-content/uploads/aal2_for_SPM12.tar.gz
AICHA	384	No	No	Yes	Built by estimation of resting-state networks, k-means clustering, homotopic regional grouping based on maximal inter-hemispheric functional correlation, and ROI labeling.	3004.333	N/A	Adaptation of AAL focused on the idea that each region in one hemisphere has a homologue in the other hemisphere	https://www.ncbi.nlm.nih.gov/pubmed/26213217	2015	Included in mricron: https://people.cas.sc.edu/rorden/mricron/index.html
Brodmann	40	Yes	Yes	No	Corticall parcellation separating regions based on cellular morphology and organization	32978.512	N/A	Brodman areas separated by gyri	http://digital.zbmed.de/zbmed/id/554966	1909	https://surfer.nmr.mgh.harvard.edu/fswiki/BrodmannAreaMaps
CAPRSC	333	Yes	No	Yes	Automatic using resting-state functional connectivity (RSFC) boundary maps	1389.09	N/A	Created using RSFC-boundary maps to define parcels that represent putative cortical areas. Focuses on the cortical surface and was created using functional MRI scans.	https://www.ncbi.nlm.nih.gov/pubmed/25316338	2016	Obtained from Freesurfer: https://sites.wustl.edu/petersenschlaggarlab/parcels-19cwpgu/
CPAC200	200	No	No	No	Created by parcellating whole brain resting-state fMRI data into spatially coherent regions of homogeneous functional connectivity.	5860.755	N/A	ROIs with anatomic homology	https://pubmed.ncbi.nlm.nih.gov/21769991/	2018	https://fcp-indi.s3.amazonaws.com/data/Projects/ABIDE_Initiative/Resources/cc200_roi_atlas.nii.gz
Desikan	70	Yes	No	No	Anatomical Landmarks based on gyri. Averaged based on majority voting	24786.857	N/A	Surface parcellation	https://www.sciencedirect.com/science/article/pii/S1053811906000437?via%3Dihub	2006	Obtained from FreeSurfer: https://surfer.nmr.mgh.harvard.edu/fswiki/CorticalParcellation
DesikanKlein	97	No	No	No	Automated labeling system that subdivided the human cerebral cortex into gyral based regions of interest.	74443.62	N/A	Gyral based parcellations	https://www.sciencedirect.com/science/article/pii/S1053811906000437?via%3Dihub	2006	Obtained from FreeSurfer: https://surfer.nmr.mgh.harvard.edu/fswiki/CorticalParcellation
Destrieux	75	Yes	No	Yes	Automatically assigned a neuroanatomical label to each location on a cortical surface model based on probabilistic information estimated from a manually labeled training set.	96280.43	MNI152	Cortical surface probabilitstic atlas	https://academic.oup.com/cercor/article/14/1/11/433466	2004	Obtained from FreeSurfer: https://surfer.nmr.mgh.harvard.edu/fswiki/CorticalParcellation
DKT	84	Yes	No	No	Automatic	85964.66	N/A	Created by using a modified Desikan protocol in order to improve segmentation and make it more suited for FreeSurfer’s classifier algorithm.	https://www.frontiersin.org/articles/10.3389/fnins.2012.00171/full	2012	Obtained from FreeSurfer: https://surfer.nmr.mgh.harvard.edu/fswiki/CorticalParcellation
Glasser	360	Yes	Yes	Yes	Semi-automated. Separated based on function, connectivity, cortical architecture, topography, and expert analysis	521.994	MNI	Cortical parcellation from multi-modal images of 210 adults in HCP	https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4990127/	2016	https://balsa.wustl.edu/file/show/3VLx
Hammersmith	83	No	No	Yes	Algorithm using prior information from 30 normal adult brain MR images, which had been manually segmented to create 30 atlases, each labeling 83 anatomical structures.	19947.72	MNI152	Automatic segmentation of young children's brains into 83 regions of interest.	https://www.sciencedirect.com/science/article/pii/S1053811907010634?via%3Dihub	2003	http://brain-development.org/brain-atlases/adult-brain-atlases/adult-brain-maximum-probability-map-hammers-mith-atlas-n30r83-in-mni-space/
HarvardOxford	48	No	Yes	Yes	Created by subdividing neocortex by topographic criteria into 48 parcellation units corresponding to the principal cerebral gyri.	21966.104	N/A	Neuroanatomic subdivisions delineated by this general segmentation generaly corresponding to natural gray matter boundaries.	https://www.sciencedirect.com/science/article/pii/S0920996405004998?via%3Dihub	2005	Included in FSL: https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Atlases
JHU	48	Yes	No	Yes	One subject manually labelled and warped to 29 other adult atlases (Large Deformation Diffeomorphic Metric Mapping)	3541.792	N/A	A small version of a larger (289 ROI) atlas composed based on parcellation of deep white matter. Split into 4 groups: Tracts in the brainstem, projection fibers, association fibers, and commisural fibers	https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2724595/	2004	https://neurovault.org/collections/264/
Juelich	103	No	No	Yes	Probabilistic atlas created by averaging multi-subject post-mortem cyto- and myelo-architectonic segmentations.	69433	N/A	N/A	https://linkinghub.elsevier.com/retrieve/pii/S1053-8119(04)00792-X	2005	https://interactive-viewer.apps.hbp.eu/?templateSelected=MNI+Colin+27&parcellationSelected=JuBrain+Cytoarchitectonic+Atlas
MICCAI	136	No	No	No	N/A	52708.26	N/A	N/A	MICCAI 2012 Workshop: https://my.vanderbilt.edu/masi/workshops/	2012	http://www.neuromorphometrics.com/2012_MICCAI_Challenge_Data.html
Princeton	49	Yes	Yes	Yes	Identified 25 topographic maps in a large population of individual subjects and transformed them into either surface- or volume-based standardized space.	1217.388	N/A	Atlas exclusively containing parcellations of the visual cortex.	https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4585523/	2015	https://scholar.princeton.edu/napl/resources
Schaefer1000	1000	No	No	No	Automatic using gwMRF	1055.685	N/A	Gradient-weighted Markov Random Fields (gwMRF) to group similar fMRI regions (dependent on # of regions specified)	http://people.csail.mit.edu/ythomas/publications/2018LocalGlobal-CerebCor.pdf	2017	https://github.com/ThomasYeoLab/CBIG/tree/master/stable_projects/brain_parcellation/Schaefer2018_LocalGlobal https://minhaskamal.github.io/DownGit/#/home?url=https://github.com/ThomasYeoLab/CBIG/tree/master/stable_projects/brain_parcellation/Schaefer2018_LocalGlobal/Parcellations
Schaefer200	200	No	No	No	Automatic using gwMRF	5278.425	N/A	Gradient-weighted Markov Random Fields (gwMRF) to group similar fMRI regions (dependent on # of regions specified)	http://people.csail.mit.edu/ythomas/publications/2018LocalGlobal-CerebCor.pdf	2017	https://github.com/ThomasYeoLab/CBIG/tree/master/stable_projects/brain_parcellation/Schaefer2018_LocalGlobal https://minhaskamal.github.io/DownGit/#/home?url=https://github.com/ThomasYeoLab/CBIG/tree/master/stable_projects/brain_parcellation/Schaefer2018_LocalGlobal/Parcellations
Schaefer300	300	No	No	No	Automatic using gwMRF	3518.95	N/A	Gradient-weighted Markov Random Fields (gwMRF) to group similar fMRI regions (dependent on # of regions specified)	http://people.csail.mit.edu/ythomas/publications/2018LocalGlobal-CerebCor.pdf	2017	https://github.com/ThomasYeoLab/CBIG/tree/master/stable_projects/brain_parcellation/Schaefer2018_LocalGlobal https://minhaskamal.github.io/DownGit/#/home?url=https://github.com/ThomasYeoLab/CBIG/tree/master/stable_projects/brain_parcellation/Schaefer2018_LocalGlobal/Parcellations
Schaefer400	400	No	No	No	Automatic using gwMRF	2639.213	N/A	Gradient-weighted Markov Random Fields (gwMRF) to group similar fMRI regions (dependent on # of regions specified)	http://people.csail.mit.edu/ythomas/publications/2018LocalGlobal-CerebCor.pdf	2017	https://github.com/ThomasYeoLab/CBIG/tree/master/stable_projects/brain_parcellation/Schaefer2018_LocalGlobal https://minhaskamal.github.io/DownGit/#/home?url=https://github.com/ThomasYeoLab/CBIG/tree/master/stable_projects/brain_parcellation/Schaefer2018_LocalGlobal/Parcellations
Slab1068	1068	No	No	No	Calculated spatially averaged time series for each of 1068 regions of interest placed in a regular 12-mm grid throughout the brain	493.719	N/A	Grid of ROI points spanning entire MNI brain volume.	https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5507181/	2017	https://umich.app.box.com/s/w46icx4bng1mw1nc3sg72t13ug5ecyib https://www.nitrc.org/projects/kessler_jama16/
Slab907	907	No	No	No	Placed 907 ROIs at regular intervals throughout the cortex	7952.68	N/A	Grid of ROI points spanning entire MNI brain volume	https://pubmed.ncbi.nlm.nih.gov/25225387/	2014	https://umich.app.box.com/s/jowv4pogzhbfevt301n8
Talairach	1105	No	Yes	Yes	Semi-automated?	1698.114	Talairach coordinates	A hierarchical atlas split into 5 leves: Hemisphere, Lobe, Gyrus, Tissue Type, and Cell Type	https://www.ncbi.nlm.nih.gov/pubmed/7008525	1980	http://www.talairach.org/
Tissue	3		No	No	N/A	609031.667	N/A		(Tissue-based segmentation: WM, GM, CSF)	2018
Yeo 17	17	Yes	No	Yes	Clustered to identify networks of functionally coupled regions	31040.294	FreeSurfer surface space	Local networks confined to sensory and motor cortices, functional connectivity followed topographic representations across adjacent areas	https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3174820/	2011	https://ftp.nmr.mgh.harvard.edu/pub/data/Yeo_JNeurophysiol11_MNI152.zip http://www.freesurfer.net/fswiki/CorticalParcellation_Yeo2011
Yeo 17 Liberal	17	Yes	No	Yes	Clustered to identify networks of functionally coupled regions	62043.118	FreeSurfer surface space	Local networks confined to sensory and motor cortices, functional connectivity followed topographic representations across adjacent areas	https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3174820/	2011	https://ftp.nmr.mgh.harvard.edu/pub/data/Yeo_JNeurophysiol11_MNI152.zip http://www.freesurfer.net/fswiki/CorticalParcellation_Yeo2011
Yeo 7	7	Yes	No	Yes	Clustered to identify networks of functionally coupled regions	75383.571	FreeSurfer surface space	Local networks confined to sensory and motor cortices, functional connectivity followed topographic representations across adjacent areas	https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3174820/	2011	https://ftp.nmr.mgh.harvard.edu/pub/data/Yeo_JNeurophysiol11_MNI152.zip http://www.freesurfer.net/fswiki/CorticalParcellation_Yeo2011
Yeo 7 Liberal	7	Yes	No	Yes	Clustered to identify networks of functionally coupled regions	150676.143	FreeSurfer surface space	Local networks confined to sensory and motor cortices, functional connectivity followed topographic representations across adjacent areas	https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3174820/	2011	https://ftp.nmr.mgh.harvard.edu/pub/data/Yeo_JNeurophysiol11_MNI152.zip http://www.freesurfer.net/fswiki/CorticalParcellation_Yeo2011
DS Family	71 - 72784	No	No	No	Grid segmentation of entire MNI brain	Variable	N/A	Grid segmentation of entire MNI brain	https://ieeexplore.ieee.org/document/6736874	2013	N/A

neuroparc's Issues

image gain

Hi!
Thank you so much for your great service,But I seem to have some problems that I need your help with,the file (Tissue_space-MNI152NLin6_res-1x1x1.nii.gz) is the sgementation label or raw file, if the former, Where do I go to get the original file?
thanks!
chenyong

can we remove the folder called "ebridge" please?

Atlas Summary

I think it would be helpful to organize the information about each of the atlases in this repo. I would like to see a table with a row for each atlas and each column in that row containing some statistics about the atlas. This would make it easy to compare information about different atlases and maybe there would be some that suit specific purposes that people could identify. Ex:

	Volume	Labels	Center	etc...
Atlas 1
Atlas 2
...

suggestion: list the network labeling version for the Schaefer parcellations.

The Schaefer parcellations should have the corresponding number of networks specified in their name or metadata, as the number of parcels is not sufficient for identification. For example, the neuroparc Schaefer400_space-MNI152NLin6_res-2x2x2.nii.gz matches Schaefer2018_Parcellations/MNI/Schaefer2018_400Parcels_7Networksorder_FSLMNI152_2mm.nii.gz, but not Schaefer2018_Parcellations/MNI/Schaefer2018_400Parcels17Networks_order_FSLMNI152_2mm.nii.gz.

thanks!

column in table for whether subcortical parcels exist in atlas

I received a few emails recently from colleagues looking for atlases that include subcortical coverage. Might be helpful to mention such features in the table...

DesikanKlein atlas out-brain values

Hi!
First, thanks for the great effort here!
I just came across that some voxels in the 1x1x1 sapce of the Desikan Atlas are out of the brain region (i.e label 1009). Maybe you should mask by the files in /mask to get rid of those.

datalad fork of the data content

I created a datalad dataset to store the data content of this repo and to get around the GIT LFS issues mentionned (#36, #35, #26)

https://gin.g-node.org/cpp-lln-lab/neuroparc_data

are Glasser, Yeo et al, Power et al) here?

can not open any atlases or nifti images using MRIcron or ITKsnap

Not sure why but I can not open any atlases or nifti templates using MRIcron or ITKsnap
I have got with ITKsnap:

Error: Unsupported or missing image file format. ITK-SNAP failed to create an ImageIO object for the image 'neuroparc-master/atlases/label/Human/AAL_space-MNI152NLin6_res-2x2x2.nii.gz' using format ''.

Region sorting is non canonical for (at least) Schaefer parcellations

When looking at the Schaefer 200 parcellation here in neuroparc, and comparing it to the canonical parcellation produced by Thomas Yeo's lab, I noticed that the regions are inconsistently sorted. For example, note the intensity values in the screenshot (shown in the header bar; first number is neuroparc label and second number is YeoLab label), and the general colour inconsistency (additive map is used, meaning that regions that have the same value in both parcellations are perfectly gray; none other than the background).

When opening it in a Python shell, I created a mapping from one to the other — there is a 1-1 correspondence, so all regions are the same size, just differently numbered. I've attached a plot to visualize this mapping, below. When looking at the metadata of the neuroparc copy, the label name is null for all regions, meaning that I cannot reverse-engineer network membership from this metadata, either.

My questions are as follows:

1. How come the regions are sorted differently? Is this intentional, or an error?
2. How come any processing happened to the canonical file? Is it documented in some way, and is it easily reversible?
3. Does this issue affect (or was this processing applied to) any other parcellations? While I only investigated this version because I stumbled across it, it limits my enthusiasm to use the other enclosed parcellations here, as I feel I would first need to find the original source to verify it, and at that point, neuroparc isn't providing any more convenience.

make BIDS extension PR

https://docs.google.com/document/d/1pWmEEY-1-WuwBPNy5tDAxVJYQ9Een4hZJM06tQZg8X4/edit?usp%3Dsharing&sa=D&ust=1537468908724000

add atlases from nilearn that we are missing

https://nilearn.github.io/modules/reference.html#module-nilearn.datasets

then make a PR into nilear for a function that fetches any of our atlases.

also discuss with them making a PR to store all the atlases in nilearn or a nilearn repo, maybe using gitlfs

Make consistent with TemplateFlow?

I am very happy to see movement on this topic. Thank you so much for putting this together. I see a couple of issues that might be good to address in the future:

1. BIDS apps like fMRIprep use Templateflow as a source of their standard templates. Unfortunately, Templateflow proscribes RAS whereas you seem to have landed on LAS ; (

2. Your templates are a slightly different FOV than Templateflow:

MNI152NLin6_res-1x1x1_T1w.nii.gz 182 x 218 x 182
Whereas the corresponding TemplateFlow image, tpl-MNI152NLin6Sym/tpl-MNI152NLin6Sym_res-01_T1w.nii.gz
is 193 x 229 x 193

I think it would help all of us if NeuroParc and TemplateFlow matched orientation and FOV.
Thanks for considering this.

-Dianne

Cannot open atlases nor metadata

AFter downloading, each file only contains github version, ID and size. Example of the AAL atlas in: https://github.com/neurodata/neuroparc/blob/master/atlases/label/Human/AAL_space-MNI152NLin6_res-1x1x1.nii.gz

version https://git-lfs.github.com/spec/v1
oid sha256:020f4a0d7bee976d910ca2c9ff19adab7f34facaa60909d1cbbccaecd83a5de1
size 29680

When this should be a compresssed nifti file.
The exact same information is contained in the cvs and json metadata files.

Cannot work out file type

Hi there!

I'm trying to run the DICE correlation between the 1x1x1 Destrieux and the 1x1x1 Yeo 7 liberal atlases in a jupyter notebook and I'm running into an issue in the fourth cell. I keep getting the following error:

I have copied the atlas names exactly (with the MNI152NLin6 and the .nii.gz) and am pretty positive that the input directory is correct. I was also able to successfully import nibabel, numpy, argparse, and matplotlib in the first cell. Let me know if there's something else I need to be doing to move forward with the correlation, thanks!

add other species atlases

• mouse
• rat
• zebrafish
• coyote
• sea lion
• some primates

AAL Labels

The AAL atlas in this database is the AAL1 Atlas with 116 parcels
But the label file of AAL has 120 parcels and these are the labels from the AAL2 atlas
Tried to upload corrected file #40

`MetaData["Number of Regions"]` is inconsistent

Contribution Criteria § File 2: Atlas Information says

Number of Regions: The number of regions defined by the atlas, not including empty space. Will be calculated if not provided

and that is accurate in the Atlas Info Summary table in all of the atlases I've inspected, but in the JSON files, many do include empty space in the count, so the number in the MetaData["Number of Regions"] field in the JSON is one more than the number in the # Regions column in the README.

Relatedly,

neuroparc/scripts/json_generation.py

Lines 165 to 175 in 5a5e746

	js_contents['rois'][str(0)] = {"label": "empty", "center":None}
	for (k, v) in csv_dict.items():
	k=int(k)
	try:
	js_contents['rois'][str(k)] = {"label": v, "center": parcel_centers[k], "size":int(size[k])}
	roi_sum=roi_sum+size[k]
	count=count+1
	except KeyError:
	js_contents['rois'][str(k)] = {"label": v, "center": None, "size": None}
	except TypeError:
	js_contents['rois'][str(k)] = {"label": v, "center": None, "size": None}

will include the region with index 0 in the count if that index is included in the label CSV.

nilearn has a bunch of standard atlases

https://nilearn.github.io/modules/reference.html#module-nilearn.datasets

1. let's make sure we have all of those
2. how are they storing their networks, maybe we want to consider storing ours the same way?
3. maybe we want to make a PR to nilearn and add a bunch of additional atlases?

@ebridge2

Add fields to all json specifications

Many of the jsons for different parcellations already have labels pre-computed, but it would be of value to ensure that all have fields for "parcel name", "Hemisphere" (L/R), "Tissue Type" (GM, WM, CSF), "Yeo7 Network" (1 -> 7) "Center of Mass" ([a,b,c], in mms), lobe pre-computed for all parcellations.

Rename "Desikan"

... Since, it has undergone some undocumented transformations relative to the original Desikan parcellation.

Could we please add to the scripts folder....

Users are going to want a set of simple API's for saving and retrieving all kinds of metadata from the .json. Otherwise, folks might not catch on?

Add "source" and "reference" field

Add a "source" field, and a "reference" field, giving the link to the source from which the atlas can originally be downloaded, and a reference for citation of the parcellation to the json specification

Installation details in README.md

Hello,

Thanks for gathering all these atlases in one place. It's very handy.

I naively attempted to download the atlases via the "Download as Zip" button on Github. This doesn't work due to the use of git-lfs. Perhaps the README.md should start with details of how to obtain the atlases via git clone and explain why?

Also, at the end of cloning, my version of git-lfs reported that two files "were not pointers, but should have been".

integrate with templateflow

templateflow/templateflow#15

consider atlas reader

http://joss.theoj.org/papers/10.21105/joss.01257

Left-right flipping possibility?

This repository has an impressive number of atlases. I took a quick look at two of these atlases that were of particular interest, and it appears that both may be flipped left-right relative to an MNI template. I looked at the Glasser and Brodmann atlases. It's not clear which MNI template is being used, so I can't be completely sure. Still neither match the 2009c asymmetric nor the original MNI152 templates. The usual petalia with the left hemisphere slightly more posterior and the right hemisphere slightly anterior seems to be backwards here. Both qform and sform codes (value=4 for MNI) and values for the sform_rows seem to be set in the NIFTI header, so the problem is not a missing header but possibly an incorrect one. I will note that we have what I think is an improved version of the MNI transformed Glasser atlas in the AFNI distribution, and soon a better Brodmann atlas too.

Missing insula in Human/Anatomical-labels-csv/Desikan.csv?

Hello,

I'm trying to draw a correspondence between the neuroparc Desikan ROIs and the MNE-python's Deiskan ROI labels (read from freesurfer lh/rh.aparc.annot files - mne.read_labels_from_annot).

It seems to me that:

1. neuroparc's Human/Anatomical-labels-csv/Desikan.csv is missing the insula L/R labels.
2. there are no corresponding labels for neuroparc's "white_matter" and "corpus_calosum" in the freesurfer list below.

I'm new to brain anatomy and parcellations, so not sure if I missed/misunderstood something! Thank you for the very helpful repo! :)

---

For reference, MNE-python reads in the following:

import mne
mne_labels = mne.read_labels_from_annot("fsaverage", parc="aparc")
mne_labels = [x.name for x in mne_labels]
mne_labels, len(mne_labels)

Reading labels from parcellation...
read 35 labels from fsaverage/label/lh.aparc.annot
read 34 labels from fsaverage/label/rh.aparc.annot
(['bankssts-lh',
'bankssts-rh',
'caudalanteriorcingulate-lh',
'caudalanteriorcingulate-rh',
'caudalmiddlefrontal-lh',
'caudalmiddlefrontal-rh',
'cuneus-lh',
'cuneus-rh',
'entorhinal-lh',
'entorhinal-rh',
'frontalpole-lh',
'frontalpole-rh',
'fusiform-lh',
'fusiform-rh',
'inferiorparietal-lh',
'inferiorparietal-rh',
'inferiortemporal-lh',
'inferiortemporal-rh',
'insula-lh',
'insula-rh',
'isthmuscingulate-lh',
'isthmuscingulate-rh',
'lateraloccipital-lh',
'lateraloccipital-rh',
'lateralorbitofrontal-lh',
'lateralorbitofrontal-rh',
'lingual-lh',
'lingual-rh',
'medialorbitofrontal-lh',
'medialorbitofrontal-rh',
'middletemporal-lh',
'middletemporal-rh',
'paracentral-lh',
'paracentral-rh',
'parahippocampal-lh',
'parahippocampal-rh',
'parsopercularis-lh',
'parsopercularis-rh',
'parsorbitalis-lh',
'parsorbitalis-rh',
'parstriangularis-lh',
'parstriangularis-rh',
'pericalcarine-lh',
'pericalcarine-rh',
'postcentral-lh',
'postcentral-rh',
'posteriorcingulate-lh',
'posteriorcingulate-rh',
'precentral-lh',
'precentral-rh',
'precuneus-lh',
'precuneus-rh',
'rostralanteriorcingulate-lh',
'rostralanteriorcingulate-rh',
'rostralmiddlefrontal-lh',
'rostralmiddlefrontal-rh',
'superiorfrontal-lh',
'superiorfrontal-rh',
'superiorparietal-lh',
'superiorparietal-rh',
'superiortemporal-lh',
'superiortemporal-rh',
'supramarginal-lh',
'supramarginal-rh',
'temporalpole-lh',
'temporalpole-rh',
'transversetemporal-lh',
'transversetemporal-rh',
'unknown-lh'],
69)

Original resources? and how were the surface-based parcellations projected to the MNI volume?

Hey,

Is there a document about where each of the parcellations was originally downloaded from? Some of the parcellations (Glasser, Schaefer etc.) were originally created on the cortical surface. How were these parcellations projected back to the MNI volume (e.g. registration method from Wu2017?

Thanks!
Ting

Mistake in HarvardOxfordcort-maxprob.csv

Hi,
Thank you so much for this great tool!
I believe there is a mistake in the HarvardOxfordcort-maxprob.csv file:
Label 7 says "Subcallosal_Cortex" when it should be "Precentral_Gyrus", as evident by the anatomy, and also here: https://neurovault.org/images/1699/

Thanks!
Roey

The below code snippets should help ensure consistent orientations and voxel dimensions!

Assuming ndmg is installed, then:

import numpy as np
import nibabel as nib
import os
import sys
import shutil

Modify image orientation:

def reorient_t1w(t1w, namer):
    cmd='fslorient -getorient ' + t1w
    orient = os.popen(cmd).read().strip('\n')
    if orient == 'NEUROLOGICAL':
        print('Reorienting derivative t1w image to RAS+ canonical...')
        # Orient t1w to std
        t1w_orig = t1w
        t1w = "{}/t1w_reor.nii.gz".format(namer.dirs['output']['prep_anat'])
        shutil.copyfile(t1w_orig, t1w)
        cmd='fslorient -forceradiological ' + t1w_orig
        os.system(cmd)
        cmd='fslreorient2std ' + t1w_orig + ' ' + t1w
        os.system(cmd)
    else:
	print('No reorientation of derivative t1w image needed...')
        t1w_orig = t1w
        t1w = "{}/t1w.nii.gz".format(namer.dirs['output']['prep_anat'])
        shutil.copyfile(t1w_orig, t1w)
    return t1w

Modify voxel resolution:

def match_target_vox_res(img_file, vox_size, namer, zoom_set, sens):
    from ndmg.utils import reg_utils as rgu
    # Check dimensions
    img = nib.load(img_file)
    hdr = img.get_header()
    zooms = hdr.get_zooms()
    if (round(abs(zooms[0]), 0), round(abs(zooms[1]), 0), round(abs(zooms[2]), 0)) is not zoom_set:
	if sens == 'dwi':
            img_file_pre = "{}/{}_pre_res.nii.gz".format(namer.dirs['output']['prep_dwi'], os.path.basename(img_file).split('.nii.gz')[0])
	elif sens == 't1w':
	    img_file_pre = "{}/{}_pre_res.nii.gz".format(namer.dirs['output']['prep_anat'], os.path.basename(img_file).split('.nii.gz')[0])
        shutil.copyfile(img_file, img_file_pre)
        if vox_size == '1mm':
            print('Reslicing image to 1mm...')
            img_file = rgu.reslice_to_xmm(img_file_pre, 1.0)
        elif vox_size == '2mm':
            print('Reslicing image to 2mm...')
            img_file = rgu.reslice_to_xmm(img_file_pre, 2.0)
    return img_file

Glasser parcellation inaccurate

I was playing around with the different parcellations and noticed that the Glasser2016 parcellation is transformed inaccurately.

I then went to the https://osf.io/67a3t/files/ website and found that the Glasser parcellation over there was fine. Probably a version mismatch.

Cross-map between atlases

Hi, thanks for this awesome project.
I was wondering if you have any tools to cross-map between atlases by region label and to get labels for each atlas based on coordinates in standardized spaces.
Thanks a lot,

Pedro

Reorganization

I would probably do "Human" at a layer above where it is now; ie, "Human/label/...", "Macaque/..." this way the templates, masks, parcellations, and transforms associated with a single species are in one place. Right now it would probably be less intuitive than it should be to separate out by species, which I imagine is what 99.9% of investigators would want.

Also, please delete the jsons called "_updated.json"; if they are correct and as desired, they should just be the jsons, not called "updated.json".

make filenames BIDS compliant

As far as I can tell most of the filenames are not BIDS valid filenames as they do not have any suffix.

The most obvious choice to align with BIDS would be to follow the filenaming templates for discrete segementations: https://bids-specification.readthedocs.io/en/stable/05-derivatives/03-imaging.html#segmentations

Additionally the labels .csv files should probably be turned into TSV files that also follow the BIDS derivatives spec

problem with opening/reading atlases in label/Human (it is similar with other .nii.gz files)

I tried fslview to open the files --> example error message
** ERROR (nifti_image_read): bad binary header read for file './neuroparc-master/atlases/Human/AAL2zourioMazoyer2002.nii.gz'

• read 130 of 348 bytes
  ** ERROR: nifti_image_open(./neuroparc-master/atlases/Human/AAL2zourioMazoyer2002): bad header info
  ERROR: failed to open file ./neuroparc-master/atlases/Human/AAL2zourioMazoyer2002
  Qt has caught an exception thrown from an event handler. Throwing
  exceptions from an event handler is not supported in Qt. You must
  reimplement QApplication::notify() and catch all exceptions there.

Failed to open file ./neuroparc-master/atlases/Human/AAL2zourioMazoyer2002

Part of FSL (build 4110)
fslview (4.0.1)

Copyright(c) 2005, University of Oxford
Dave Flitney

Usage:
fslview [-m 3d|ortho|lightbox] [-l lutname] [-b low,hi]
[ [-l lutname] [-b low,hi] ] ...
fslview -m ortho,lightbox filtered_func_data thresh_zstat1 -t 0.5 thresh_zstat2 -l "Cool" -t 0.5

Optional arguments (You may optionally specify one or more of):
-V,--verbose switch on diagnostic messages
-h,--help display this message
-m,--mode Initial viewer mode. Comma separated list of: 3d; single, ortho; lightbox

Per-image options

Usage:
image [-l GreyScale] [-t 0.1] [-b 2.3,6]
-l,--lut Lookup table name. As per GUI, one of: Greyscale;
"Red-Yellow"; "Blue-Lightblue"; Red; Green;
Blue; Yellow; Pink; Hot; Cool; Copper, etc.
-b,--bricon Initial bricon range, e.g., 2.3,6
-t,--trans Initial transparency, e.g., 0.2

Also tried to open with register:

** ERROR (nifti_image_read): bad binary header read for file 'yeo-17-liberal_space-MNI152NLin6_res-2x2x2.nii.gz'

• read 130 of 348 bytes
  ** ERROR: nifti_image_open(yeo-17-liberal_space-MNI152NLin6_res-2x2x2.nii.gz): bad header info

and then when I tried gunzip says the
gzip: yeo-7-liberal_space-MNI152NLin6_res-1x1x1.nii.gz: not in gzip format for all the files

Identify which parcellations are canonical

Related to #48

So that users can identify which templates are canonical, versus those that have been transformed in someway, it would be important to include this information. In an MVP, this can just be adding a dictionary key of "is different", but in a perfect world, there should also be details on the transformations themselves.

CPAC200_space-MNI152NLin6_res-2x2x2.nii.gz appears to be misaligned in MNI space

Hello!

It seems that the CPAC200 atlas is slightly misaligned in MNI space, as shown in the attached screenshot. A version which appears correct is here: http://preprocessed-connectomes-project.org/abide/Pipelines.html