Summary of what happened:
Hi all,
I have several questions regarding fieldmaps and distortion correction in fmriprep. I’d first like to report what I’m observing. I ask whether this is intended behaviorl; If not, I’d love to know how to troubleshoot.
- observation 1: Number of fieldmaps for sourcedata and fmriprep output differ - is this intended behavior?
- observation 2: Number of fmriprep-generated-fmaps differ from subject to subject
- observation 3: json IntendedFor field differs between sourcedata and fmriprep
Thank you in advance for your help.
Observation 1: Number of fieldmaps for sourcedata and fmriprep output differ - is this intended behavior?
- In this example, I grab the same participant, comparing sourcedata vs. fmriprep fmap numbers
- sourcedata has 2 fieldmaps
- fmriprep generates 3 fieldmaps for the same sub/session
# source data: {SOURCEDIR}/sub-0061/ses-04/fmap
sub-0061_ses-04_acq-mb8_dir-ap_run-01_epi.json sub-0061_ses-04_acq-mb8_dir-ap_run-02_epi.nii.gz sub-0061_ses-04_acq-mb8_dir-pa_run-02_epi.json
sub-0061_ses-04_acq-mb8_dir-ap_run-01_epi.nii.gz sub-0061_ses-04_acq-mb8_dir-pa_run-01_epi.json sub-0061_ses-04_acq-mb8_dir-pa_run-02_epi.nii.gz
sub-0061_ses-04_acq-mb8_dir-ap_run-02_epi.json sub-0061_ses-04_acq-mb8_dir-pa_run-01_epi.nii.gz
# fmriprep data: {FMRIPREPDIR}/sub-0061/ses-04/fmap
sub-0061_ses-04_acq-mb8_fmapid-auto00002_desc-coeff_fieldmap.nii.gz sub-0061_ses-04_acq-mb8_fmapid-auto00003_desc-preproc_fieldmap.json
sub-0061_ses-04_acq-mb8_fmapid-auto00002_desc-epi_fieldmap.nii.gz sub-0061_ses-04_acq-mb8_fmapid-auto00003_desc-preproc_fieldmap.nii.gz
sub-0061_ses-04_acq-mb8_fmapid-auto00002_desc-preproc_fieldmap.json sub-0061_ses-04_acq-mb8_fmapid-auto00004_desc-coeff_fieldmap.nii.gz
sub-0061_ses-04_acq-mb8_fmapid-auto00002_desc-preproc_fieldmap.nii.gz sub-0061_ses-04_acq-mb8_fmapid-auto00004_desc-epi_fieldmap.nii.gz
sub-0061_ses-04_acq-mb8_fmapid-auto00003_desc-coeff_fieldmap.nii.gz sub-0061_ses-04_acq-mb8_fmapid-auto00004_desc-preproc_fieldmap.json
sub-0061_ses-04_acq-mb8_fmapid-auto00003_desc-epi_fieldmap.nii.gz sub-0061_ses-04_acq-mb8_fmapid-auto00004_desc-preproc_fieldmap.nii.gz
Observation 2: Number of fmriprep-generated-fmaps differ from subject to subject
- In this example, I compare two participants and check fmriprep-generated fieldmap numbers
- sub-0061 has 3 fieldmaps generated via fmriprep
- (Note that sub-0061 sourcedata had 2 fieldmaps; listed above in #observation 1)
- sub-0031 has 2 fieldmaps generated via fmriprep
- Why do participants have different number of fmriprep-generated fmaps?
# sub-0061 fmriprep generated fmap number: 3
sub-0061_ses-04_acq-mb8_fmapid-auto00002_desc-coeff_fieldmap.nii.gz sub-0061_ses-04_acq-mb8_fmapid-auto00003_desc-preproc_fieldmap.json
sub-0061_ses-04_acq-mb8_fmapid-auto00002_desc-epi_fieldmap.nii.gz sub-0061_ses-04_acq-mb8_fmapid-auto00003_desc-preproc_fieldmap.nii.gz
sub-0061_ses-04_acq-mb8_fmapid-auto00002_desc-preproc_fieldmap.json sub-0061_ses-04_acq-mb8_fmapid-auto00004_desc-coeff_fieldmap.nii.gz
sub-0061_ses-04_acq-mb8_fmapid-auto00002_desc-preproc_fieldmap.nii.gz sub-0061_ses-04_acq-mb8_fmapid-auto00004_desc-epi_fieldmap.nii.gz
sub-0061_ses-04_acq-mb8_fmapid-auto00003_desc-coeff_fieldmap.nii.gz sub-0061_ses-04_acq-mb8_fmapid-auto00004_desc-preproc_fieldmap.json
sub-0061_ses-04_acq-mb8_fmapid-auto00003_desc-epi_fieldmap.nii.gz sub-0061_ses-04_acq-mb8_fmapid-auto00004_desc-preproc_fieldmap.nii.gz
# sub-0031 fmriprep generated fmap number: 2
sub-0031_ses-04_acq-mb8_fmapid-auto00001_desc-coeff_fieldmap.nii.gz sub-0031_ses-04_acq-mb8_fmapid-auto00002_desc-coeff_fieldmap.nii.gz
sub-0031_ses-04_acq-mb8_fmapid-auto00001_desc-epi_fieldmap.nii.gz sub-0031_ses-04_acq-mb8_fmapid-auto00002_desc-epi_fieldmap.nii.gz
sub-0031_ses-04_acq-mb8_fmapid-auto00001_desc-preproc_fieldmap.json sub-0031_ses-04_acq-mb8_fmapid-auto00002_desc-preproc_fieldmap.json
sub-0031_ses-04_acq-mb8_fmapid-auto00001_desc-preproc_fieldmap.nii.gz sub-0031_ses-04_acq-mb8_fmapid-auto00002_desc-preproc_fieldmap.nii.gz
Observation 3: json IntendedFor field differs between sourcedata and fmriprep
- In this example, I check the intendedFor fields for sourcedata and fmriprep for an identical sub/ses
Sourcedata json IntendedFor
- The IntendedFor field in the sourcedata lists all the corresponding runs with identical shim settings. This is automatically populated via Heudiconv
- Notice that the sourcedata json only includes task A and B, as intended.
fmriprep json Intended For
- However, fmriprep-generated IntendedFor field includes task A, B, and C (all of the tasks in ses-04)
- It’s also listing all of the fmaps as RawSources (instead of just one)
# sourcedata fmap json for sub-0061, ses-04 first fieldmap.
# NOTE: there are two fieldmaps in this specific session
"IntendedFor": [
"ses-04/func/sub-0061_ses-04_task-A_acq-mb8_run-01_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-A_acq-mb8_run-02_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-B_acq-mb8_run-01_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-B_acq-mb8_run-02_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-B_acq-mb8_run-03_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-B_acq-mb8_run-04_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-B_acq-mb8_run-05_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-B_acq-mb8_run-06_bold.nii.gz"
],...
# fmriprep fmap json
{
"AnatomicalReference": "sub-0061_ses-04_acq-mb8_fmapid-auto00002_desc-epi_fieldmap.nii.gz",
"AssociatedCoefficients": [
"sub-0061_ses-04_acq-mb8_fmapid-auto00002_desc-coeff_fieldmap.nii.gz"
],
"B0FieldIdentifier": "auto_00002",
"IntendedFor": [
"ses-04/func/sub-0061_ses-04_task-C_acq-mb8_run-01_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-C_acq-mb8_run-02_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-A_acq-mb8_run-01_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-A_acq-mb8_run-02_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-B_acq-mb8_run-01_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-B_acq-mb8_run-02_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-B_acq-mb8_run-03_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-B_acq-mb8_run-04_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-B_acq-mb8_run-05_bold.nii.gz",
"ses-04/func/sub-0061_ses-04_task-B_acq-mb8_run-06_bold.nii.gz"
],
"RawSources": [
"/bids/sub-0061/ses-04/fmap/sub-0061_ses-04_acq-mb8_dir-ap_run-01_epi.nii.gz",
"/bids/sub-0061/ses-04/fmap/sub-0061_ses-04_acq-mb8_dir-ap_run-02_epi.nii.gz",
"/bids/sub-0061/ses-04/fmap/sub-0061_ses-04_acq-mb8_dir-pa_run-01_epi.nii.gz",
"/bids/sub-0061/ses-04/fmap/sub-0061_ses-04_acq-mb8_dir-pa_run-02_epi.nii.gz"
],
"Units": "Hz"
}
sub-0061_ses-04_acq-mb8_fmapid-auto00002_desc-preproc_fieldmap.json (END)
Command used (and if a helper script was used, a link to the helper script or the command generated):
singularity run \
--cleanenv \
-B "${DCS}":/main \
-B "${BIDS}":/bids \
-B "${OUT}":/out \
-B "${WORK}":/scratch \
-B "${ANAT}":/anat \
/users/${USER}/fmriprep_21.0.2.sif \
/bids /out/fmriprep participant -w /scratch \
"${flags[@]}" \
--notrack -vv --cifti-output 91k \
--mem_mb 252000 --nprocs 56 --fs-subjects-dir /anat/sourcedata/freesurfer \
--fs-license-file /main/fmriprep/tools/license.txt \
--skip-bids-validation \
--random-seed 2022 --skull-strip-fixed-seed \
--bids-database-dir /main/inputs/bids_db
Version:
fmriprep 21.0.2
Environment (Docker, Singularity, custom installation):
Singularity
Data formatted according to a validatable standard? Please provide the output of the validator:
Relevant log outputs (up to 20 lines):
Screenshots / relevant information:
Here’s my scanner sequence. Hope this helps to see the order of runs in relation to the fieldmaps
anat-scout_ses-04
fmap-epi_dir-ap_acq-mb8_run-01
fmap-epi_dir-pa_acq-mb8_run-01
func_run-01_task-A_acq-ap-mb8
func_run-02_task-A_acq-ap-mb8
func_run-01_task-B_acq-ap-mb8
func_run-02_task-B_acq-ap-mb8
func_run-03_task-B_acq-ap-mb8
func_run-04_task-B_acq-ap-mb8
func_run-05_task-B_acq-ap-mb8
func_run-06_task-B_acq-ap-mb8
anat-scout_ses-04
fmap-epi_dir-ap_acq-mb8_run-02
fmap-epi_dir-pa_acq-mb8_run-02
func_run-01_task-C_acq-ap-mb8
func_run-02_task-C_acq-ap-mb8