Fmriprep: How fmriprep deals with multiple T1 ?

[tib-sim]

Hi,
I am wondering how fmriprep deals with mutliple T1 ?
In my case we did only one T1 acquisition but we applied a "retro-reconstruction" on this image by removing the "pre-scan normalization filter". This resulted in a second T1, that we aim to use enventually for structural analysis. Once I converted my data in BIDS, I tried to keep all images, thus the first T1 was defined with the suffix "run-001" and the second with the suffix "run-002".
I am wondering if this can caused issue during the preprocessing done by fmriprep ?
Best,
Simon

[bpinsard]

Hi Simon,
The default behavior of fmriprep is to average all T1, with freesurfer mri_robust_template.
In your case I would say that it is not optimal to average the same acquisition with 2 reconstruction options.
You can use the option --bids-filter-file to have more specific way to select which file you want to use.
Here is the doc on how to use it.
https://fmriprep.org/en/latest/faq.html#how-do-i-select-only-certain-files-to-be-input-to-fmriprep

Best.

[emdupre]

Just to +1 @bpinsard's suggestion !

One other note: run is probably not the way I would distinguish these two files in BIDS. You might want to look instead at the rec entitiy: https://bids-specification.readthedocs.io/en/stable/04-modality-specific-files/01-magnetic-resonance-imaging-data.html#the-rec-entity

[tib-sim]

Thank you for your replies and help !
Even if I use the rec entity, I guess I will have to use in any case the --bids-filter-file option by specifying which T1 to select in the JSON file. I guess rec does not allow to avoid the averaging of all T1 by fmriprep ?
Best,
Simon

[mgxd]

Nope, fmriprep does not make any inclusion / exclusion assumptions based on the rec entity.