T1s and fMRIs in MNI space?
I am starting to work with the preprocessed full T1s, rfMRIs and tfMRI and would like to know if the images are in the MNI space.
My intention is to apply some functional atlas on the preprocessed data, which means that the data I need is probably the one found in filtered_func_data_clean for fMRIs and in T1_brain for T1s. From the documentation, it seems a little bit confusing to decipher whether or not images are in the MNI space that many atlases need to mask the data appropriately for example.
My question comes up based on this paragraph from the documentation: “The EPI unwarping is a combined step that also includes alignment to the T1, though the unwarped data is written out in native (unwarped) fMRI space (and the transform to T1 space written out separately). This T1 alignment is carried out by FLIRT, with a final BBR cost function [Greve and Fischl, 2009]. After the fMRI GDC unwarping, a final FLIRT realignment to T1 is applied, to take into account any shifts resulting from the GDC unwarping. The previously described transform from T1 space to standard MNI space is utilised when fMRI data is needed in standard space.”
It is then my understanding that the fMRIs are aligned to the T1s transformed to MNI space, however the information about T1s says that the standard-space brain mask is then back transformed into the space of the T1 and applied to the T1, generating the brain-extracted T1s. For me that suggests that the T1s are not in MNI and only a middle step, in which the T1s were aligned, was used to align the fMRIs. Is it then that fMRIs are in MNI but T1s are not?
Any clarification or guidance would be highly appreciated
Comments
4 comments
Hi Sebastian,
The T1_brain.nii.gz and rfMRI.nii.gz files are not in MNI, but are aligned to each other.
As a consequence the warps provided in the transforms subfolder of the T1 directory should be sufficient to warp both the T1_brain.nii.gz and rfMRI.nii.gz files to MNI.
An MNI aligned T1_brain is provided in the T1_brain_to_MNI.nii.gz file for each subject.
Fidel Alfaro Almagro - this is my understanding, please could you provide any corrections/additional clarifications if required.
Hi all,
Both T1_brain.nii.gz and filtered_func_data_clean.nii.gz are both in their respective native spaces, but they are not aligned to each other. You can see that by checking the size and resolution of those images:
There is an affine matrix to take the fMRI images to T1 space: fMRI/rfMRI.ica/reg/example_func2highres.mat
This can be used as the premat in applywarp in combination with the T1_to_MNI warp to take the fMRI to MNI space.
Cheers
Thanks Fidel!
Thanks for the feedback. That was really useful!
Please sign in to leave a comment.