FIX: ensure fieldwarp == identity during resampling of processed multi-echo data

[markushs]

Changes proposed in this pull request

Likely fixes issues reported in #2727

Documentation that should be reviewed

No

[effigies]

@oesteban Could I get a second pair of eyes on this? I think it's right, but I want to be extra careful.

@markushs Are you able to test this patch on your installation? Do you need any guidance?

[markushs]

@markushs Are you able to test this patch on your installation? Do you need any guidance?

@effigies I'll try to bind an edited version of base.py into my singularity image. Should work, but as this image reflects "vanilla" 21.0.1 I won't be able to test the patch against newly merged stuff.

[markushs]

My tests appear successful.
*_space-T1w*_bold.nii.gz-files are now identical to volumes in native functional "scanner"-space transformed to space-T1w using the fmriprep-generated *_from-scanner_to-T1w_mode-image_xfm.txt.

command.txt in bold_t1_trans_wf/bold_to_t1w_transform looks correct for all volumes, with 2x --transform identity (i added the backslashes for readibility):

Not sure what the empty _dpop function in the now deleted lines was doing, but everything looks good without, so maybe consider also getting rid of the function definition?

[effigies]

Thanks for confirming!

[oesteban]

LGTM (opened #2732 as a reminder for myself)