Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add template: MNI152NLin6Sym #6

Closed
oesteban opened this issue Jan 8, 2019 · 16 comments

Comments

Projects
None yet
6 participants
@oesteban
Copy link
Contributor

commented Jan 8, 2019

[comes from https://github.com/poldracklab/smriprep/issues/11]

http://nist.mni.mcgill.ca/?p=858

As per the original discussion, MNI152NLin6ASym was proposed. However, I can't find such template on the MNI website. @feilong, could you confirm that you are interested in the symmetric template (i.e., MNI152NLin6Sym)?

@feilong

This comment has been minimized.

Copy link

commented Jan 8, 2019

Hi @oesteban , my purpose then was to make it possible to apply atlases available in FSL to fmriprep outputs. I cannot recall the details, but from the records it seems it should be the asymmetric version.

@oesteban

This comment has been minimized.

Copy link
Contributor Author

commented Jan 8, 2019

Could you provide us with the link?

@feilong

This comment has been minimized.

Copy link

commented Jan 8, 2019

I don't know where the source is, but it's available as part of FSL: http://neuro.debian.net/pkgs/fsl-mni152-templates.html

@oesteban

This comment has been minimized.

Copy link
Contributor Author

commented Jan 8, 2019

Okay, as per their documentation (https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Atlases) FSL's atlas corresponds to the "152 nonlinear 6th generation" atlas. So I guess that the only existing version is the Symmetric one.

@feilong

This comment has been minimized.

Copy link

commented Jan 8, 2019

Based on the original discussion (poldracklab/fmriprep#337) and the reference (http://www.lead-dbs.org/?p=1241) the asymmetric version was used, but that version isn't available from the official website...

@jnvandermeer

This comment has been minimized.

Copy link

commented Jan 16, 2019

I tried to look for the asymmetric version of MNI152 -6- Nlim:
http://packages.bic.mni.mcgill.ca/mni-models/icbm152/mni_icbm152_nl_VI_minc2.zip -- BUT - these consist of only symmetric images.

We have the same issue now that we wish to have processed bold images using the MNI152Nlim6Asym space (to be consistent with other work), and are looking to apply a deformation field to the images as a final processing step. It appears that the difference between version 6 and version 2009c of these templates is a position shift and a stretch in A-P direction.

Does anyone has a good link for the MNI152Nlim6Asym in nifti format? Or is it as easy as taking the T1 image from the fsl package (in /fsl/5.0/data/standard/MNI152_T1_1mm_brain.nii.gz)?

Thanks for your time
Johan

@oesteban

This comment has been minimized.

Copy link
Contributor Author

commented Jan 16, 2019

Let's page @vfonov, he probably knows why MNI is not releasing the asymmetric version of the 2006 nonlinear templates.

@vfonov

This comment has been minimized.

Copy link

commented Jan 16, 2019

in 2006 we did not create asymmetric version.

@oesteban

This comment has been minimized.

Copy link
Contributor Author

commented Jan 16, 2019

I see, I'll ping the FSL list to check the provenance of the asymmetric version. Thanks very much Vladimir.

@vfonov

This comment has been minimized.

Copy link

commented Jan 16, 2019

Maybe @andrewjanke was involved in that

@andrewjanke

This comment has been minimized.

Copy link

commented Jan 17, 2019

Yes, I made the FSL model. It was made using a nigh on identical method to Vlad's. We were in the same room at the time under Louis and were pretty much writing code to do this at the same time. Perhaps mine predated his slightly. My Perl version was volgenmodel (https://github.com/andrewjanke/volgenmodel) vlads was in Ruby IIRC. make_a_model?

There is also a NiPypy version of volgenmodel: https://github.com/carlohamalainen/volgenmodel-nipype

I would still have the original files somewhere, a nonlinear match between them was made at some point in time. Both to the nl6 and to the previous FSL models.

vlad you may be able to find these in ~rotor/data on the MNI systems.

@vfonov

This comment has been minimized.

Copy link

commented Jan 17, 2019

yes, my code is in https://github.com/vfonov/build_average_model
And the python version is a module model in the pipeline code: https://github.com/vfonov/nist_mni_pipelines

@oesteban

This comment has been minimized.

Copy link
Contributor Author

commented Jan 17, 2019

@andrewjanke would you be willing to provide the original file set for us to add it as a template to this repository? As you can see, we are trying to provide a regular interface to all (possible) templates (and associated atlases and metadata).

BTW @vfonov, we would love to hear any feedback from your expert perspective (same for Andrew). Please let us know any thoughts if there are some.

@oesteban

This comment has been minimized.

Copy link
Contributor Author

commented Mar 28, 2019

Hi @andrewjanke, what is/are the most relevant references that should be present when using the asymmetric template you created (i.e. when using FSL's template)?

Metadata associated to the template - https://github.com/templateflow/tpl-MNI152NLin6Asym/blob/0b6662e5ebd650acd2bdd6d2d2d90fbcf9ff8728/template_description.json#L1-L32

@oesteban

This comment has been minimized.

Copy link
Contributor Author

commented Mar 29, 2019

I've just finished adding both templates and indexing them within the superdataset.

@oesteban oesteban closed this Mar 29, 2019

@andrewjanke

This comment has been minimized.

Copy link

commented Mar 29, 2019

oesteban added a commit to templateflow/tpl-MNI152NLin6Asym that referenced this issue Apr 1, 2019

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
You can’t perform that action at this time.