I tried implementing the T2* driven coregistration using the --t2s-coreg flag (fMRIPrep version 1.5.8). It finishes with no errors, but looking through the output file, I noticed this line: “No multiecho BOLD images found for T2* coregistration. Using standard EPI-T1 coregistration.”
In order for T2* coregistration to work, does the BOLD data have to be multiecho? And would fMRIPrep detect multiecho data automatically?
Yes, to use the T2* coregistration, the BOLD data must be multi-echo. fMRIPrep would automatically detect the multi-echo data, assuming it’s in BIDS format.
Any idea what’s the --t2s-coreg equivalent for fMRIPrepv20.2.1? Or does fMRIPrep now automatically detect a multi-echo dataset and output an optimally combined image?
There is no equivalent anymore. We dropped the --t2s-coreg approach from fMRIPrep because many folks were finding that coregistration was not improved when using the T2* map. However, that doesn’t impact anything other than the coregistration.
Exactly! Unless you feed in individual echoes with --echo-idx, it will combine echoes automatically.