n4_correct fails for datasets with multi-volume sbrefs

[effigies]

What happened?

Looking into poldracklab/tacc-openneuro#68, I found that N4BiasFieldCorrection fails on 4D images, producing a non-zero exit code but no output. Adding verbosity:

Exception caught: 
itk::ExceptionObject (0x5ffe4b476700)
Location: "unknown" 
File: /home/chris/mambaforge/envs/fmriprep/include/ITK-5.3/itkImageToImageFilter.hxx
Line: 218
Description: ITK ERROR: N4BiasFieldCorrectionImageFilter(0x5ffe4b45b410): Inputs do not occupy the same physical space! 
InputImage Origin: [-9.2676048e+01, -9.8007235e+01, 6.1100900e+01], InputImageMaskImage Origin: [-9.2631440e+01, -9.7714432e+01, 6.8613268e+01]
	Tolerance: 9.6191998e-06
InputImage Direction: -0.0000000e+00 9.9992417e-01 -1.2315087e-02
9.9700160e-01 -9.5299441e-04 -7.7374607e-02
-7.7380978e-02 -1.2278075e-02 -9.9692603e-01
, InputImageMaskImage Direction: 1.0000000e+00 0.0000000e+00 0.0000000e+00
0.0000000e+00 1.0000000e+00 0.0000000e+00
0.0000000e+00 0.0000000e+00 1.0000000e+00

	Tolerance: 1.0000000e-06

By running fslroi $SBREF ./sbref0.nii.gz 0 1, I was able to run it without problems.

I think the correct approach is to run the initial boldref workflow on the sbref before passing it to enhance_and_skullstrip_wf. Need to check whether to that will work with single volume sbrefs or we need to make it conditional.

What command did you use?

NIPYPE_CONFIG_DIR=$PWD/code fmriprep sourcedata/raw . participant --notrack --skip-bids-validation -vv --sloppy --participant-label 114 -w /data/scratch/$( basename $PWD ) --bids-filter-file $PWD/code/bids-filter.json --level minimal

What version of fMRIPrep are you running?

24.0.0

How are you running fMRIPrep?

Local installation ("bare-metal")

Is your data BIDS valid?

Yes

Are you reusing any previously computed results?

No

Please copy and paste any relevant log output.

(fmriprep) church ➜  n4_correct cat command.txt 
N4BiasFieldCorrection --bspline-fitting [ 200 ] -d 3 --input-image /data/testing/fmriprep-24-ds004489/sourcedata/raw/sub-114/ses-1/func/sub-114_ses-1_task-catLoc_run-1_sbref.nii.gz --output sub-114_ses-1_task-catLoc_run-1_sbref_corrected.nii.gz -r --weight-image /data/scratch/fmriprep-24-ds004489/fmriprep_24_0_wf/sub_114_wf/bold_ses_1_task_catLoc_run_1_wf/bold_fit_wf/enhance_and_skullstrip_bold_wf/fix_header/tpl-MNI152NLin2009cAsym_res-01_label-brain_probseg_trans_fixhdr.nii.gz
(fmriprep) church ➜  n4_correct $( cat command.txt ) || echo $?
1
(fmriprep) church ➜  n4_correct $( cat command.txt ) -v 1 || echo $?

Running N4 for 3-dimensional images.

Mask not read.  Using the entire image as the mask.

Original intensity range:  [0, 4912]

***********************************************************
Warning:  Your input image contains nonpositive values
which could cause failure or problematic results.  A
possible workaround would be to:
   1. rescale your image to positive values e.g., [10,100].
   2. run N4 on your rescaled image.
   3. (optional) rescale the N4 output to the original
      intensity range.
***********************************************************

Specified spline distance: 200mm
  original image size:  [84, 84, 51]
  padded image size:  [84, 84, 84]
  number of control points:  [4, 4, 4]

Exception caught: 
itk::ExceptionObject (0x5ffe4b476700)
Location: "unknown" 
File: /home/chris/mambaforge/envs/fmriprep/include/ITK-5.3/itkImageToImageFilter.hxx
Line: 218
Description: ITK ERROR: N4BiasFieldCorrectionImageFilter(0x5ffe4b45b410): Inputs do not occupy the same physical space! 
InputImage Origin: [-9.2676048e+01, -9.8007235e+01, 6.1100900e+01], InputImageMaskImage Origin: [-9.2631440e+01, -9.7714432e+01, 6.8613268e+01]
	Tolerance: 9.6191998e-06
InputImage Direction: -0.0000000e+00 9.9992417e-01 -1.2315087e-02
9.9700160e-01 -9.5299441e-04 -7.7374607e-02
-7.7380978e-02 -1.2278075e-02 -9.9692603e-01
, InputImageMaskImage Direction: 1.0000000e+00 0.0000000e+00 0.0000000e+00
0.0000000e+00 1.0000000e+00 0.0000000e+00
0.0000000e+00 0.0000000e+00 1.0000000e+00

	Tolerance: 1.0000000e-06



1

Additional information / screenshots

No response