PennLINC / qsiprep

Preprocessing of diffusion MRI
http://qsiprep.readthedocs.io
BSD 3-Clause "New" or "Revised" License
138 stars 55 forks source link

TotalReadoutTime and --ignore fieldmaps #198

Closed ins0mniac2 closed 8 months ago

ins0mniac2 commented 3 years ago

If I understand correctly, TotalReadoutTime is needed for distortion correction. If I don't have it and choose to forego SDC with topup or fieldmaps and therefore "--ignore fieldmaps", qsiprep will still throw an error

File "/usr/local/miniconda/lib/python3.7/site-packages/qsiprep/interfaces/fmap.py", line 1005, in topup_inputs_from_4d_file spec_lookup[unique_dwi] = spec_line % spec['TotalReadoutTime'] TypeError: must be real number, not NoneType

bwinsto2 commented 3 years ago

Not sure if this is useful for you, but I had to calculate TotalReadoutTime manually. This link is helpful.

mattcieslak commented 3 years ago

qsiprep uses TotalReadoutTime for defining distortion groups, which determine which scans can be concatenated. It's a good idea to find out what this value is regardless of whether you're using distortion correction or not, but if you're really sure that it doesn't matter you could put a dummy number in there like 0.1.

ins0mniac2 commented 3 years ago

Somehow didn't get a notification of replies, thanks both of you guys for responding. That link is indeed helpful but still not enough is available in the legacy data for us to calculate it. @mattcieslak : it's not that it doesn't matter, but I want to be able to process without it. From your suggestion, sounds like I can add a dummy value and still be able to process with "--ignore fieldmaps" so that qsiprep will not throw an error, but will continue to process without SDC.

mattcieslak commented 8 months ago

Please reopen if this issue comes up again, I believe it worked?