Susceptibility distortion correction: None

,
1

Running pilots using fmriprep-docker. Preprocessed data shows no SDC:

Susceptibility distortion correction: None

Created BIDS folders w/ associated .nii/.json using dcm2bids version 2.1.6. When running BIDS validator, receiving no errors. For task pilot MID, the end of BIDS file includes the IntendedFor:

"IntendedFor": [
        "func/sub-01_ses-1_task-mid_run-01_bold.nii.gz",
        "func/sub-01_ses-1_task-mid_run-02_bold.nii.gz"
    ]
}

Complete .json for AP.json file is provided: sub-31_ses-1_dir-AP_epi-mid.txt (2.7 KB)

Below is the folder structure for a given subject (truncating excess tasks/ses). Unclear as to why the fieldmaps are not recognized/utilized.

Sub has two sessions; func has three tasks (mid [2] runs; faces [1] run; resting [1] run)

sub-01
├── ses-1
│   ├── anat
│   │   ├── sub-01_ses-1_T1w.json
│   │   ├── sub-01_ses-1_T1w.nii.gz
│   │   ├── sub-01_ses-1_inplaneT1.json
│   │   └── sub-01_ses-1_inplaneT1.nii.gz
│   ├── fmap
│   │   ├── sub-01_ses-1_dir-AP_epi-faces.json
│   │   ├── sub-01_ses-1_dir-AP_epi-faces.nii.gz
│   │   ├── sub-01_ses-1_dir-AP_epi-mid.json
│   │   ├── sub-01_ses-1_dir-AP_epi-mid.nii.gz
│   │   ├── sub-01_ses-1_dir-AP_epi-resting.json
│   │   ├── sub-01_ses-1_dir-AP_epi-resting.nii.gz
│   │   ├── .....
│   └── func
│       ├── sub-01_ses-1_task-mid_run-01_bold.json
│       ├── sub-01_ses-1_task-mid_run-01_bold.nii.gz
│       ├── sub-01_ses-1_task-mid_run-01_events.tsv
│       ├── sub-01_ses-1_task-mid_run-01_recording-cardiac_physio.json
│       ├── sub-01_ses-1_task-mid_run-01_recording-cardiac_physio.tsv.gz
│       ├── sub-01_ses-1_task-mid_run-01_recording-respiratory_physio.json
│       ├── sub-01_ses-1_task-mid_run-01_recording-respiratory_physio.tsv.gz
│       ├── sub-01_ses-1_task-mid_run-02_bold.json
│       ├── sub-01_ses-1_task-mid_run-02_bold.nii.gz
│       ├── sub-01_ses-1_task-mid_run-02_events.tsv
│       ├── sub-01_ses-1_task-mid_run-02_recording-cardiac_physio.json
│       ├── ...
└── ses-2
    ├── ....```
2

Hi,

Please refer to the bids specification for the files in fmap/ folder. The suffix should be epi in your case.

1 Like
3

I’m replying to this thread because the solution above hasn’t worked for me.

I’m running fMRIPrep v25.0.0 locally, which keeps returning output without distortion correction (Susceptibility distortion correction: None). The data set is BIDS-valid. Here is the folder structure for one subject:

sub-010001
|–– ses-02
|   |–– anat
|   |   |–– sub-010001_ses-02_acq-mp2rage_T1w.json
|   |   |__ sub-010001_ses-02_acq-mp2rage_T1w.nii.gz
|   |–– fmap
|   |   |–– sub-010001_ses-02_acq-GEfmap_run-01_magnitude1.json
|   |   |–– sub-010001_ses-02_acq-GEfmap_run-02_magnitude1.json
|   |   |–– sub-010001_ses-02_acq-GEfmap_run-01_magnitude1.nii.gz
|   |   |–– sub-010001_ses-02_acq-GEfmap_run-02_magnitude1.nii.gz
|   |   |–– sub-010001_ses-02_acq-GEfmap_run-01_magnitude2.json
|   |   |–– sub-010001_ses-02_acq-GEfmap_run-02_magnitude2.json
|   |   |–– sub-010001_ses-02_acq-GEfmap_run-01_magnitude2.nii.gz
|   |   |–– sub-010001_ses-02_acq-GEfmap_run-02_magnitude2.nii.gz
|   |   |–– sub-010001_ses-02_acq-GEfmap_run-01_phasediff.json
|   |   |–– sub-010001_ses-02_acq-GEfmap_run-02_phasediff.json
|   |   |–– sub-010001_ses-02_acq-GEfmap_run-01_phasediff.nii.gz
|   |   |__ sub-010001_ses-02_acq-GEfmap_run-02_phasediff.nii.gz
|   |–– func
|   |   |–– sub-010001_ses-02_task-rest_acq-AP_run-01_bold.json
|   |   |–– sub-010001_ses-02_task-rest_acq-PA_run-01_bold.json
|   |   |–– sub-010001_ses-02_task-rest_acq-AP_run-01_bold.nii.gz
|   |   |–– sub-010001_ses-02_task-rest_acq-PA_run-01_bold.nii.gz
|   |   |–– sub-010001_ses-02_task-rest_acq-AP_run-02_bold.json
|   |   |–– sub-010001_ses-02_task-rest_acq-PA_run-02_bold.json
|   |   |–– sub-010001_ses-02_task-rest_acq-AP_run-02_bold.nii.gz
|   |   |__ sub-010001_ses-02_task-rest_acq-PA_run-02_bold.nii.gz

As far as I can tell, the fmap json files have the correct fields (including “IntendedFor” fields pointing to the correct subject-relative path). Here is the json file for the phasediff run-01 file above:

{
    "RepetitionTime": 0.68,
    "EchoTime1": 0.00519,
    "EchoTime2": 0.00765,
    "MagneticFieldStrength": 3,
    "FlipAngle": 60,
    "PhaseEncodingDirection": "i-",
    "ManufacturersModelName": "Verio",
    "IntendedFor": [
        "ses-02/func/sub-010001_ses-02_task-rest_acq-AP_run-01_bold.nii.gz",
        "ses-02/func/sub-010001_ses-02_task-rest_acq-PA_run-01_bold.nii.gz"
    ],
    "ImageType": [
        "ORIGINAL",
        "PRIMARY",
        "P",
        "ND"
    ],
    "Manufacturer": "Siemens",
    "SkullStripped": false,
    "B0PhaseIdentifier": "phasediff_fmap0"
}

Here is the fmriprep command I’ve been running. Notably, I’m skipping skull-stripping because I’ve already skull-stripped the T1 anatomical volumes using SynthStrip.

fmriprep \
	${root_dir}/nordic \
	${root_dir}/fmriprep \
	participant \
	--participant_label ${subj} \
	--fs-no-reconall \
	--skull-strip-t1w skip \
	--output-spaces MNI152NLin2009cAsym \
	--clean-workdir \
	--work-dir ${root_dir}/fmriprep/work

Is there some minor formatting error that is preventing fMRIPrep from running distortion correction? Could it be a problem with using the pre-skull-stripped anatomical? Thanks in advance for the help getting distortion correction to run correctly.

4

Hi @Phil, and welcome to neurostars!

If not using a container, you should use a container.

It looks like you also tried to specify the B0FieldIdentifier/Source but included a typo (Phase instead of Field).

If you specify both the IntendedFor and B0Field* methods, fmriprep will ignore the IntendedFor. Perhaps first try making sure you correctly specify your B0FieldSource/B0FieldIdentifier pairs.

In your phasediff jsons, have "B0FieldIdentifier": "whatever_you_want_to_name_fmap"

In your BOLD jsons, include "B0FieldSource": "whatever_you_named_corresponding_fmap"

Also, it is recommended to not include --fs-no-reconall.

The BIDS validator output would also be useful in seeing if there are other errors.

Best,
Steven