Questions regarding FSL topup for fMRI data

i’ve managed to run the following successfully

fslmerge -t mergedfmaps.nii.gz AP_SEfmap.nii.gz PA_SEfmap.nii.gz
topup --imain=mergedfmaps.nii.gz --datain=acq_param.txt --config=b02b0.cnf --out=topup
applytopup --imain=AP.nii.gz,PA.nii.gz --inindex=1,4 --datatin=…/acq_param.txt --topup=topup --out=corr.nii.gz

AP.nii.gz and PA.nii.gz are basically two rsfMRI runs within the same session, each containing 647 volumes. According to topup/ExampleTopupFollowedByApplytopup - FslWiki , the output ‘corr.nii.gz’ is basically combined from AP.nii.gz and PA.nii.gz. Since these are fMRI volumes, it doesn’t quite make sense to combine both volumes into a single volume?

is it possible to obtain fieldmap corrected versions of both AP.nii.gz and PA.nii.gz ?

It is not entirely unreasonable to combine them, as long as you factor that in later (e.g adding a column in your design matrix to indicate which run a volume is from).

If you want to separate the runs again you can do that too. Several software packages should be able to do this (e.g. fslsplit Followed by fslmerge).

To separate the concatenated volume, if you know how many volumes are in AP and PA, you can directly use fslroi in one step to separate the two series again after susceptibility correction.

Or use applytop separately on the AP and PA series perhaps?

the output volume consist of 647 volumes— the same number in the AP or PA.nii.gz, so it doesnt make much sense to split it to obtain the two series.

but anyway i sorta figured it out. Yes, i think applytop should be applied separately on the AP and PA series, according to biac:analysis:topup_correction [Brain Imaging & Analysis Center]