fmriprep: Sophisticated PEPOLAR schemes are unsupported

What happened?

Dear fmriprep team,

Thanks for the update! I am looking forward to testing the changes! (also see https://twitter.com/DalenbergJr/status/1471392667481300994).

I have PA AP scans but PEPOLAR is not working in v21 while it ran in 20.2.5.

The console output tells me: “Sophisticated PEPOLAR schemes are unsupported” but I do not understand why it’s “sophisticated”. It’s a regular PA vs AP setup.

I am looking forward to use this implementation because the previous PEPOLAR implementation failed on half my multi-echo data set and I had to use syn-sdc (and sometimes even had to remove phase encoding direction info) to get the distortion correction right.

-Jelle

What command did you use?

docker run -ti --rm \
    -v /share/nemo/data/source/neuroimaging/MRI/test/:/data:ro \
    -v /share/nemo/data/derivatives/MRI/test/:/out \
    -v /home/data/software/freesurfer/license.txt:/opt/freesurfer/license.txt \
    -v /home/data/tmp/scratch/:/scratch \
    nipreps/fmriprep:21.0.0 /data /out participant --participant-label sub-p1197 \
        --fs-no-reconall --nprocs 6 --fs-license-file /opt/freesurfer/license.txt -t hands2 -w /scratch

What version of fMRIPrep are you running?

21.0.0

How are you running fMRIPrep?

Docker

Is your data BIDS valid?

Yes

Are you reusing any previously computed results?

Please copy and paste any relevant log output.

Setting-up fieldmap "auto_00000" (EstimatorType.PEPOLAR) with <sub-p1197_dir-PA_run-1_epi.nii.gz, sub-p1197_task-rest1_echo-1_bold.nii.gz, sub-p1197_task-rest1_echo-2_bold.nii.gz, sub-p1197_task-rest1_echo-3_bold.nii.gz, sub-p1197_task-hands1_echo-1_bold.nii.gz, sub-p1197_task-hands1_echo-2_bold.nii.gz, sub-p1197_task-hands1_echo-3_bold.nii.gz>
Process Process-2:
Traceback (most recent call last):
  File "/opt/conda/lib/python3.8/multiprocessing/process.py", line 315, in _bootstrap
    self.run()
  File "/opt/conda/lib/python3.8/multiprocessing/process.py", line 108, in run
    self._target(*self._args, **self._kwargs)
  File "/opt/conda/lib/python3.8/site-packages/fmriprep/cli/workflow.py", line 118, in build_workflow
    retval["workflow"] = init_fmriprep_wf()
  File "/opt/conda/lib/python3.8/site-packages/fmriprep/workflows/base.py", line 85, in init_fmriprep_wf
    single_subject_wf = init_single_subject_wf(subject_id)
  File "/opt/conda/lib/python3.8/site-packages/fmriprep/workflows/base.py", line 452, in init_single_subject_wf
    raise NotImplementedError(
NotImplementedError: Sophisticated PEPOLAR schemes are unsupported.

Additional information / screenshots

If you need more info, let me know!

About this issue

Original URL
State: closed
Created 3 years ago
Comments: 24 (11 by maintainers)

Most upvoted comments

21.0.1 is released.

effigies on Jan 24, 2022

Hi, thanks for following up! We had a call about this this morning, and are going to do two things:

Restore the old interpretation of IntendedFor to make sure that users can get that behavior, if they desire. (Long-term, probably 21.0.1 or 21.0.2.)
Clarify the metadata necessary to achieve what you actually want to achieve. (In this post)

Case 1: Blip-up/blip-down fieldmaps + BOLD file

sub-01/
  func/
    sub-01_task-rest_dir-PA_bold.json
    sub-01_task-rest_dir-PA_bold.nii.gz
  fmap/
    sub-01_dir-AP_epi.json
    sub-01_dir-AP_epi.nii.gz
    sub-01_dir-PA_epi.json
    sub-01_dir-PA_epi.nii.gz

In this case, both sub-01_dir-AP_epi.json and sub-01_dir-PA_epi.json would include something like

  "B0FieldIdentifier": "pepolar",

and sub-01_task-rest_dir-PA_bold.json would include

  "B0FieldSource": "pepolar",

Case 2: BOLD file + reverse-blip fieldmap

sub-01/
  func/
    sub-01_task-rest_dir-PA_bold.json
    sub-01_task-rest_dir-PA_bold.nii.gz
  fmap/
    sub-01_dir-AP_epi.json
    sub-01_dir-AP_epi.nii.gz

To achieve the expected result where the first few volumes of BOLD are combined with the EPI to estimate the fieldmap, then we need to add B0FieldIdentifier to both sub-01_dir-AP_epi.json and sub-01_task-rest_dir-PA_bold.json:

  "B0FieldIdentifier": "pepolar",

so sub-01_task-rest_dir-PA_bold.json would look include

  "B0FieldIdentifier": "pepolar",
  "B0FieldSource": "pepolar",

Case 3: Independently correcting multiple BOLD files with one reverse-blip fieldmap (Legacy behavior)

sub-01/
  func/
    sub-01_task-rest_dir-PA_run-1_bold.json
    sub-01_task-rest_dir-PA_run-1_bold.nii.gz
    sub-01_task-rest_dir-PA_run-2_bold.json
    sub-01_task-rest_dir-PA_run-2_bold.nii.gz
  fmap/
    sub-01_dir-AP_epi.json
    sub-01_dir-AP_epi.nii.gz

Because we’re independently correcting, we need to use multiple identifiers for sub-01_dir-AP_epi.json:

  "B0FieldIdentifier": ["pepolar1", "pepolar2"],

And then each BOLD file would include itself as part of one of these identifiers:

  "B0FieldIdentifier": "pepolar1",
  "B0FieldSource": "pepolar1",

  "B0FieldIdentifier": "pepolar2",
  "B0FieldSource": "pepolar2",

Case 4: Estimating a single fieldmap for all BOLD files, with one reverse-blip fieldmap

sub-01/
  func/
    sub-01_task-rest_dir-PA_run-1_bold.json
    sub-01_task-rest_dir-PA_run-1_bold.nii.gz
    sub-01_task-rest_dir-PA_run-2_bold.json
    sub-01_task-rest_dir-PA_run-2_bold.nii.gz
  fmap/
    sub-01_dir-AP_epi.json
    sub-01_dir-AP_epi.nii.gz

In this case, we can use the first volumes of all PA BOLD runs to find a single fieldmap that we will use to correct all BOLD runs. In this case, we would use a single identifier for sub-01_dir-AP_epi.json:

  "B0FieldIdentifier": "pepolar",

And then each BOLD file would include itself as part of this identifier and be corrected by that identifier

  "B0FieldIdentifier": "pepolar",
  "B0FieldSource": "pepolar",

Note that these might produce errors, we haven’t really tested these cases yet. But we’ve agreed (please check the above @oesteban) that we should be supporting these cases.

effigies on Jan 7, 2022