Parcellation for MNI152NLin2009cAsym?



I am looking for a subcortical atlas that can be used with the MNI152NLin2009cAsym template (default of fmriprep). I found a parcellation file from the fmriprep image named /niworkflows_data/mni_icbm152_nlin_asym_09c/1mm_parc.nii.gz, and I guess it was converted from the Harvard Oxford Atlas in FSL based on the scripts from the same folder.

I have two questions:

  1. Is that parcellation file also in the MNI152NLin2009cAsym template space (instead of the MNI152 6th used by FSL)?
  2. Is there any associated file that can map numbers in that file to label names?




Did you find the atlas for MNI152Nlin2009cAsym? I am looking for a parcelled atlas for the symetric version. Please any help would be really appreciated.



Hi all,

We are working on this issue right now. Please check on the development of our templateflow project ( We will be adding all the templates available to fmriprep there, including mappings of segmentations, parcellations and surfaces between them. I’ll be posting a preprint ASAP.

Regarding the two questions from @feilong (sorry for the huge delay in answering, this is probably useless to you at this point but will leave this here for people asking about it):

  1. No, that parcellation corresponds to the MNI152 6th generation. We only regridded it to match MNI152NLin2009cAsym.
  2. No, but as you mention that was done using the Harvard Oxford Atlas and the labels were mapped as in the scripts you already accessed.

With templateflow we will be addressing both two questions (which are extremely relevant, that must be said). Hopefully I can post updates here soon.


I’m not sure if I understand-
is there any problem using the relevant harvard atlases from nilearn for data preprocessed with fmriprep?

A bit puzzled after reading this discussion.



Hi @brai,

This has been a longstanding problem in neuroimaging (

The problem is that Harvard atlases were produced with the MNI152 Linear version of the MNI templates. Although the MNI152Nlin2009cAsym is well aligned with the linear version, there will always be inconsistencies between atlases if they would have defined using the MNI152Nlin2009cAsym template as well.

In summary, by applying an atlas defined w.r.t. the Linear version of MNI152 on data aligned to MNI152Nlin2009cAsym you are dismissing some accuracy errors derived from the differences between templates.

Then you would think, why didn’t you use MNI152 linear as the registration target in the first place? The answer is that then you would be introducing similar accuracy errors but on your registration step. We considered that it was more important to have the best alignment possible to the standard space, assuming that the current lack of atlases defined on MNI152Nlin2009cAsym would be addressed at some point in time.

For these reasons, we are working on an easy way of transferring atlas information through different templates. That would be one of the objectives of the project I mentioned above (templateflow).

Hopefully, this addresses your concerns



Many thanks for your helpful answer!

Wondering when will semantic segmentation will take place (over each subject t1 space) instead of normalization (at least for ROI based analysis…)

split this topic #7

6 posts were split to a new topic: Availability of the Harvard-Oxford atlas in MNI152NLin2009cAsym template


After almost 2 years, I can give a proper solution to this question.

  1. We have just uploaded two parcellations (Harvard-Oxford and Schaefer2018) to TemplateFlow, in their original FSL MNI space (or MNI152NLin6Asym) and corresponding resamplings in MNI152NLin2009cAsym. The transform I used to perform this mapping was estimated with the TemplateFlow registration framework and it is also available via TemplateFlow.

  2. The associated files are under the MNI152NLin6Asym template, although I’ll try to generate more metadata for the Schaefer2018 atlas during a few more days.