I have run fMRIPrep successfully on 30 patients. Then I tried running the same fMRIPrep command on 30 control subjects, but the ‘func’ folder ended up missing from the fMRIPrep output directory. I.e. I only got the ‘anat’ and ‘fmap’ folders as output.
The patients and the control participants were all scanned in the same scanner and using the same resting-state sequence. They are named according to the BIDS standard, and the .nii.gz files seem correct. However, the JSON files belonging to the functional data of the control participants are shorter than the ones from the patients.
I understand that attempting to re-convert the DICOMs would be the best way to approach this in case this is the step that caused incomplete JSON files.
Nevertheless, my question is: in theory, would it be fine to copy a “working” JSON file from one of the (successfully preprocessed) patients and use this for preprocessing of the controls? Or would this mess things up?
I have now started running preprocessing of one control participant with a “borrowed” JSON file hoping this was the reason for the missing ‘func’ folder. However, because of a lot of time pressure, I need to ask another question before I see the final results of the processing:
According to this last log file it already looks like resting-state data is being processed in some way. This was not the case for the earlier log files (with the shorter JSONs). However it has some warnings about an “Outdated cache” both for anatomical, field map, and functional data.
I wouldn’t worry about it an outdated cache, it just means something has to be recalculated instead of reused from a previous run of fMRIprep. As long as it finishes without error you should be good!
