fMRIPrep coregister on longitudinal T2

Summary of what happened: my data structure is as follows:

ses-00(baseline): T1, T2, and resting-state

ses-01 - ses-04 (exam session): only T2 and resting-state.

What I want to achieve is: For each exam session, the BOLD data should first be registered to the T2-weighted image acquired during the same session. The session-specific T2w image should then be rigidly registered to the baseline (ses-00) T2w image, which is aligned to the baseline T1w anatomical reference. Finally, the baseline T1w image should be normalized to MNI space. These transformations should be concatenated and applied to the BOLD data in a single resampling step whenever possible to minimize interpolation. I also want to incorporate those registration steps with fmriprep too, but it seems fmriprep only include T2 when running freesurfer. Therefore, I would like to ask if it is possible to achieve the pipeline I proposed above and integrate with fmriprep. Another question I have is that if I run fmriprep with freesure, will the fmriprep output (not the reconstructed freesurfer output) registered to T2?

Command used (and if a helper script was used, a link to the helper script or the command generated):

PASTE CODE HERE

Version: fmriprep-25.2.5

Environment (Docker, Singularity / Apptainer, custom installation): Apptainer

Data formatted according to a validatable standard? Please provide the output of the validator: yes

PASTE VALIDATOR OUTPUT HERE

Relevant log outputs (up to 20 lines):

PASTE LOG OUTPUT HERE

Screenshots / relevant information:


Hi @ZHOPOK_W,

By default, if a T2w image is available, it is used for BOLD-to-Anat registration.

You can use the --subject-anatomical-reference first-lex flag to get everything aligned to the first session, which I believe is default behavior anyway.

Yes, that behavior is default in fMRIPrep.

Everything will be ultimately be registered to the T1w image. I believe fMRIPrep should also output the T2w images registered to that same reference as well.

Best,
Steven

Hi Steven,

Thank you so much for your reply! It is very helpful. However, I would like to confirm that since for exam sessions (ses-01 - ses04) I only have T2, with no T1, when fMRIPrep processes ses-01 BOLD, does bold2anat-init=t2w use the ses-01 T2w specifically, while ultimately registering to the ses-00-based anatomical reference? And for --subject-anatomical-reference first-lex, can i use it to get every exam session aligned to the first session T2 and T1? Thank you so much!

Hi @ZHOPOK_W

If you only have one T1w, then everything from all sessions will be aligned to that image.

Best,
Steven

Ok, thank you so much!
Sorry that I have so many question, I just wanted to double check that for bold2anat-init = t2w, it will register the BOLD images to session specifically T2w, right? For example, for ses-01 BOLD, it will be registered to ses-01 T2w and for ses-02 BOLD, it will be registered to ses-02 T2w.
Thank you so much!

Hi @ZHOPOK_W,

Yes, the t2 from a session should be used to register that session’s BOLD.

Best,
Steven

1 Like