10.4. India Brain Template (IBT)

10.4.1. Overview

Here we present the India Brain Templates (IBTs). This is a set of anatomical group templates and atlases from n=466 subjects from many states around India, scanned at several different sites (all 3T scanners).

There are five sets of templates and atlases, covering a lifespan range from 6-60 years, in the following cohort groupings:

  • C1: 6-11 yrs (late childhood), n=28, 46% female

  • C2: 12-18 yrs (adolescence), n=106, 47% female

  • C3: 19-25 yrs (young adulthood), n=181, 41% female

  • C4: 26-40 yrs (adulthood), n=89, 43% female

  • C5: 41-60 yrs (late adulthood), n=62, 44% female

For more details on demographics and acquisition, please see the publication, below.

Contents

For each of the above cohorts, there are multiple template and atlas datasets.

Each template is a T1-weighted (T1w) anatomical volume. There is both a group average template and a “typical” single subject version. The group average template was through multiple iterations of alignment among all the cohort subjects, with increasing refinement (via rigid, then linear affine, and then several nonlinear warping stages). This procedure was performed using the new make_template_daskpy tool in AFNI, which utilizes Python task scheduler Dask to coordinate the jobs within and between levels, managing the resources efficiently whether on a cluster or a desktop computer. The “typical” subject version of the template is the individual anatomical aligned to the group mean template with the minimal cost function value (i.e., the brain that appears to have the best match to the final mean template), providing a particularly high tissue-contrast version of the template.

Each atlas was derived from FreeSurfer parcellations. There are atlases made from each of the “2000” and “2009” parcellations from standard FreeSurfer output. Each of those is presented in two forms: a maximum probability map (MPM), which corresponds to the group average template, and the “typical” subject atlas, which corresponds to the “typical” subject template.

An example of the contents in the IBT download for each cohort is provided here:

Set of data in IBT download (the filenames are just provided for the C1 cohort, but there is the corresponding set of files for each)

IBT contents

Description

ibt_C1_template_mean.nii.gz

group mean template

ibt_C1_template_mean_mask.nii.gz

binary mask of group mean template

ibt_C1_template_typ.nii.gz

“typical” subject template

ibt_C1_template_typ_mask.nii.gz

binary mask of “typical” subject template

ibt_C1_atlas_mpm_fs2000.nii.gz

maximum probability map (MPM) atlas from FreeSurfer’s “2000” parcellation; corresponds with group mean template

ibt_C1_atlas_mpm_fs2009.nii.gz

maximum probability map (MPM) atlas from FreeSurfer’s “2009” parcellation; corresponds with group mean template

ibt_C1_atlas_typ_fs2000.nii.gz

single-subject atlas from FreeSurfer’s “2000” parcellation; corresponds with “typical” template

ibt_C1_atlas_typ_fs2009.nii.gz

single-subject atlas from FreeSurfer’s “2009” parcellation; corresponds with “typical” template

supplementary_C1/

a directory of supplementary datasets. 1) Tissue-based maps of subsets of each atlas: GM, GM ROIs (i.e., not including unspecified cortical GM, useful for tracking or correlation matrices), WM, CSF, Ventricles, “other” (such as “optic chiasm”), and “unknown”. 2) Segmentation maps: 1=GM, 2=WM, 3=CSF+Vent, 4=other+unknown.


Example images

Here are some example images from the IBT download, using just the C3 cohort (equivalent images exist for each of the five cohorts):

Axial montages (from top to bottom): group mean template, “typical” subject template, and edges of group mean template over the “typical” subject template

../_images/img_00_C3_mean.axi.png
../_images/img_00_C3_typ.axi.png
../_images/img_00_C3_edges_typ_mean.axi.jpg

The script used to make these images with ``@chauffeur_afni`` is available here: do_00_view_ibt_templates.tcsh.


MPM “2000” atlas overlaid translucently on group mean template

../_images/img_01_C3_mean_mpm_fs2000.axi.png
../_images/img_01_C3_mean_mpm_fs2000.sag.png
../_images/img_01_C3_mean_mpm_fs2000.cor.png

Individual “2000” atlas overlaid translucently on “typical” template

../_images/img_01_C3_typ_typ_fs2000.axi.png
../_images/img_01_C3_typ_typ_fs2000.sag.png
../_images/img_01_C3_typ_typ_fs2000.cor.png

Supplementary data set: ROI-like GM parcels from individual “2000” atlas overlaid opaquely on “typical” template

../_images/img_01_C3_typ_typ_atl_fs2000_gmrois.axi.png
../_images/img_01_C3_typ_typ_atl_fs2000_gmrois.sag.png
../_images/img_01_C3_typ_typ_atl_fs2000_gmrois.cor.png

Supplementary data set: Tissue map (4 types) from “2000” FS parcellation overlaid opaquely on “group mean” template

../_images/img_01_C3_mean_mpm_seg_fs2000.axi.png
../_images/img_01_C3_mean_mpm_seg_fs2000.sag.png
../_images/img_01_C3_mean_mpm_seg_fs2000.cor.png

The script used to make these images with ``@chauffeur_afni`` is available here: do_01_view_ibt_atlases.tcsh.

Download the IBT datasets

You can download and unpack these in any of the following ways:

  • (the AFNI way) copy+paste:

    @Install_IBT_DATASETS
    
  • (the plain Linux-y terminal way) copy+paste:

    curl -O https://afni.nimh.nih.gov/pub/dist/tgz/IBT_DATASETS.tgz
    tar -xvf IBT_DATASETS.tgz
    
  • (the mouseclick+ way) click on this link,
    ... and then unpack the zipped directory by either clicking on it or using the above tar command.

Citation/questions

If you make use of the IBTs (template) or IBTAs (atlases) in your research, please cite:

Holla B, Taylor PA, Glen DR, Lee JA, Vaidya N, Mehta UM, Venkatasubramanian G, Pal P, Saini J, Rao NP, Ahuja C, Kuriyan R, Krishna M, Basu D, Kalyanram K, Chakrabarti A, Orfanos DP, Barker GJ, Cox RW, Schumann G, Bharath RD, Benegal V (2020). A series of five population-specific Indian brain templates and atlases spanning ages 6 to 60 years. Hum Brain Mapp (accepted).
For questions, comments and/or suggestions, contact:
Bharath Holla (hollabharath@gmail.com)
Rose Dawn Bharath (drrosedawnbharath@gmail.com)
Daniel Glen (glend@mail.nih.gov)
Paul Taylor (paul.taylor@nih.gov)