Add Multiresolution Intrinsic Segmentation Template (MIST)

[htwangtw]

I would like to add the Multiresolution Intrinsic Segmentation Template (MIST) to templateflow.
I have one question related to the templated used when generating this atlas.
The atlas is shipped as nifti files at 3mm isotropic voxels in the ICBM2009b space. Currently templateflow only has ICBM2009c.
Should I add ICBM2009b to templateflow first and then add MIST, or should I resample the atlas to ICBM2009c?

Please advice on the best way to move forward!

[oesteban]

Hi @htwangtw

Sorry for the slow response. Fortunately, in this case it played in your favor - I have recently added ICBM2009b (both symmetric and asymmetric versions) to templateflow, so your first question has been addressed :)

• templateflow/tpl-MNI152NLin2009bSym
• templateflow/tpl-MNI152NLin2009bAsym

Then, regarding the MIST (atlas), I think this is an excellent addition to templateflow, thanks for putting it forward!

Before we go ahead, may I ask @pbellec what is the actual template MIST is referenced to? More specifically, is it the Symmetric or the Asymmetric version?

[htwangtw]

To the best of my knowledge of the paper, the data most likely is registered to the symmetric template.
The data used in the paper was processed with NIAK and the default template is MNI152NLin2009bSym

https://github.com/SIMEXP/niak/blob/2984db211a834324bd1c2b0b8ece7e520ba8e23d/bricks/t1_processing/niak_brick_t1_preprocess.m#L14-L20

@pbellec will have to confirm