Behaviour of fmap in fmriprep with 2 sessions

Hello,
I have one specific question related to longitudinal field maps. I ran fMRIPrep for one subject who has ses-1 and ses-6. While both T1W and BOLD sessions are correctly identified as two sessions (in the html report), the process incorrectly identified only one field map (fmap). I checked all parameters in the JSON files, and all of them indicate correspondence for ses-1 and ses-6.

When I examine the outputs, I see a “fmap” folder at the same level as the ses-1 and ses-6 outputs. Inside the fmap folder, there are different estimations, and the JSON file within the folder appears to be a merged fmap (as shown in the image). This is not the behavior I am expecting because I want the process to identify each fmap scan for the corresponding session. Does anyone know how to solve this?

Thank you for your assistance!

Best regards,
Patricio

Hi @PatricioSolis-Urra and welcome to Neurostars!

Please share the following to help us help you:

  1. command used to run fmriprep
  2. version of fMRIprep you are running
  3. whether your data are bids valid
  4. your fmap JSON files and the output of tree for a single subject’s raw data directory.

Best,
Steven

Hi @Steven,
Thanks for the quick response. Here the answers:

  1. Im running in neurodesk with this command:
  2. Neurodesk have now v23.0.0.
  3. Yes.
  4. This is the tree, and I share in this drive link with the fmap json files (MRI_FMAP – Google Drive).

Thanks a lot!
Best,
Patricio

Hi @PatricioSolis-Urra,

Thanks for sharing! I noticed your intendedfor fields are using the BIDS URI (that is, the bids::sub-xx beginning). While this is theoretically correct it is not yet supported by fMRIPrep. Instead, remove that prefix, so each entry in your IntendedFor should begin with ses-.

Best,
Steven

Thanks a lot @Steven.
I will try this and I will inform you if this works

Best,
Patricio

Hi @PatricioSolis-Urra,

Sounds good! I also noticed that you are also specifying B0FieldSource, which is another way of associating fmaps-to-BOLD. I don’t know what fMRIPrep’s behavior is when both the B0FieldSource and IntendedFor methods are used. So if it still isn’t working, try temporarily removing the B0FieldSource field so fMRIPrep has to use IntendedFor.

Best,
Steven