Purpose of scanner variable?

[jdkent]

Summary

The different filename treatment of GE versus SE subjects shouldn't be necessary here

Actual behavior

fieldmaps are named differently depending on scanner type

Expected behavior

fieldmaps should be named the same independent of scanner type see section 8.9 page 24 of the BIDS spec

Perhaps I need more context on how the fieldmaps are different between scanners, but otherwise I think the change should come from making the datasets BIDS compliant.

[tbweng]

Main rationale for ${scanner} variable was to set the ${dwellTime}, since they are different between scanners. e.g., for GE, extract dwellTime from *rest_bold_info.txt, since they were recon'd with GE_dcm_to_nii.sh. For SE, set to known value (note: for BIKE_ATRAIN SE only). TODO: incorporate better way to set ${dwellTime}.

RE fmap naming, my interpretation of the BIDS manual was that the naming differed by how the fmaps were acquired; e.g., SE follows 8.9.1 Case 1, whereas GE follows 8.9.3 Case 3.

[jdkent]

Thanks! dwelltime then should be set in a sidecar json file as per bids standard as EffectiveEchoSpacing with the number specified in the units of seconds.

As for the naming, if we trust the sub-<label>_ses-<label>_fieldmap_prepped.nii.gz for the SE scans to represent "showing the field inhomogeneity in each voxel" then we should remove the magnitude and phasediff files and then rename sub-<label>_ses-<label>_fieldmap_prepped.nii.gz to sub-<label>_ses-<label>_fieldmap.nii.gz and just have the json sub-<label>_ses-<label>_fieldmap_prepped.nii.gz to sub-<label>_ses-<label>_fieldmap.json

If the work has already been done to make the SE fieldmaps "prepped" I think we should just keep the prepped ones in the BIDS dataset and remove the superfluous files. Then we are confident that SE fits under the definition of case 8.9.3 Case 3: A single, real fieldmap image (showing the field inhomogeneity in each voxel)

Now the GE fieldmaps, I think I may have specified mine wrong in PACR, so at this juncture I think I need an explanation of what we got off from the scanner for both GE and SE.

I know we don't need all those files in the fmap directory, but honestly the fieldmap stuff is still confusing me, how comfortable are you with phasediff, magnitude, and voxel-wise inhomogeneity and how each of those images look? @mwvoss do you know what each of those images should look like so I can know what kind of fieldmap data we are working with?