XCP 0.14.1 fails with different image orientation

[JohannesWiesner]

Summary

Not sure if this is an issue or a feature. One of my resting state runs has different orientations:

Normal run:

• InPlanePhaseEncodingDirectionDICOM: "COL"
• PhaseEncodingDirection: "j-"
• Orientation text: Tra>Cor(...)

Swapped-looking run:

• InPlanePhaseEncodingDirectionDICOM: "ROW"
• PhaseEncodingDirection: "i"
• Orientation text: Cor>Tra(...)

fmirprep 25.2.4 had no trouble preprocessing it, but xcpd 0.14.1 fails with:

260223-11:23:11,474 nipype.workflow WARNING:
	 [Node] Error on "_copy_atlas4" (/zi/home/johannes.wiesner/work/slurm/project_dmn_dispersion_new/workdirs/xcpd/xcp_d_0_14_wf/sub_PIAMJ5O5J_ses_PEZ02_ses_PEZ02_wf/load_atlases_wf/copy_atlas/mapflow/_copy_atlas4)
260223-11:23:11,916 nipype.workflow ERROR:
	 Node _copy_atlas0 failed to run on host zislhcn0111.zi.local.
260223-11:23:11,930 nipype.workflow ERROR:
	 Saving crash info to /zi/flstorage/group_csp/analyses/johannes.wiesner/projects/project_dmn_dispersion/data/pez/xcpd_0.14.1/sub-PIAMJ5O5J/log/20260223-111112_e73cfc75-e81e-470a-8cd9-1039da132690/crash-20260223-112311-johannes.wiesner-_copy_atlas0-c2bbbc04-6c31-418f-83a4-f4d1b0743d52.txt
Traceback (most recent call last):
  File "/usr/local/miniconda/lib/python3.10/site-packages/nipype/pipeline/plugins/multiproc.py", line 67, in run_node
    result["result"] = node.run(updatehash=updatehash)
  File "/usr/local/miniconda/lib/python3.10/site-packages/nipype/pipeline/engine/nodes.py", line 525, in run
    result = self._run_interface(execute=True)
  File "/usr/local/miniconda/lib/python3.10/site-packages/nipype/pipeline/engine/nodes.py", line 643, in _run_interface
    return self._run_command(execute)
  File "/usr/local/miniconda/lib/python3.10/site-packages/nipype/pipeline/engine/nodes.py", line 769, in _run_command
    raise NodeExecutionError(msg)
nipype.pipeline.engine.nodes.NodeExecutionError: Exception raised while executing Node _copy_atlas0.

Traceback:
	Traceback (most recent call last):
	  File "/usr/local/miniconda/lib/python3.10/site-packages/nipype/interfaces/base/core.py", line 401, in run
	    runtime = self._run_interface(runtime)
	  File "/usr/local/miniconda/lib/python3.10/site-packages/xcp_d/interfaces/bids.py", line 290, in _run_interface
	    raise ValueError(
	ValueError: Existing 'Glasser' atlas affine (/zi/flstorage/group_csp/analyses/johannes.wiesner/projects/project_dmn_dispersion/data/pez/xcpd_0.14.1/atlases/atlas-Glasser/atlas-Glasser_space-MNI152NLin2009cAsym_dseg.nii.gz) is different from the input file affine (/zi/home/johannes.wiesner/work/slurm/project_dmn_dispersion_new/workdirs/xcpd/xcp_d_0_14_wf/sub_PIAMJ5O5J_ses_PEZ02_ses_PEZ02_wf/load_atlases_wf/warp_atlases_to_bold_space/mapflow/_warp_atlases_to_bold_space0/atlas-Glasser_space-MNI152NLin6Asym_res-01_dseg.nii.gz).


260223-11:23:11,948 nipype.workflow ERROR:
	 Node _copy_atlas1 failed to run on host zislhcn0111.zi.local.
260223-11:23:11,953 nipype.workflow ERROR:
	 Saving crash info to /zi/flstorage/group_csp/analyses/johannes.wiesner/projects/project_dmn_dispersion/data/pez/xcpd_0.14.1/sub-PIAMJ5O5J/log/20260223-111112_e73cfc75-e81e-470a-8cd9-1039da132690/crash-20260223-112311-johannes.wiesner-_copy_atlas1-a28a1323-558f-40a2-8974-5f3eb1d1abb9.txt
Traceback (most recent call last):
  File "/usr/local/miniconda/lib/python3.10/site-packages/nipype/pipeline/plugins/multiproc.py", line 67, in run_node
    result["result"] = node.run(updatehash=updatehash)
  File "/usr/local/miniconda/lib/python3.10/site-packages/nipype/pipeline/engine/nodes.py", line 525, in run
    result = self._run_interface(execute=True)
  File "/usr/local/miniconda/lib/python3.10/site-packages/nipype/pipeline/engine/nodes.py", line 643, in _run_interface
    return self._run_command(execute)
  File "/usr/local/miniconda/lib/python3.10/site-packages/nipype/pipeline/engine/nodes.py", line 769, in _run_command
    raise NodeExecutionError(msg)
nipype.pipeline.engine.nodes.NodeExecutionError: Exception raised while executing Node _copy_atlas1.

Traceback:
	Traceback (most recent call last):
	  File "/usr/local/miniconda/lib/python3.10/site-packages/nipype/interfaces/base/core.py", line 401, in run
	    runtime = self._run_interface(runtime)
	  File "/usr/local/miniconda/lib/python3.10/site-packages/xcp_d/interfaces/bids.py", line 290, in _run_interface
	    raise ValueError(
	ValueError: Existing 'Gordon' atlas affine (/zi/flstorage/group_csp/analyses/johannes.wiesner/projects/project_dmn_dispersion/data/pez/xcpd_0.14.1/atlases/atlas-Gordon/atlas-Gordon_space-MNI152NLin2009cAsym_dseg.nii.gz) is different from the input file affine (/zi/home/johannes.wiesner/work/slurm/project_dmn_dispersion_new/workdirs/xcpd/xcp_d_0_14_wf/sub_PIAMJ5O5J_ses_PEZ02_ses_PEZ02_wf/load_atlases_wf/warp_atlases_to_bold_space/mapflow/_warp_atlases_to_bold_space1/atlas-Gordon_space-MNI152NLin6Asym_res-01_dseg.nii.gz).

Additional details

• xcp_d version: 0.14.1
• Docker version:
• Apptainer version: singularity-ce version 4.1.4-nobl

Normal run:

Swapped:

Full log file:

xcp_d-0.14.1_sub-PIAMJ5O5J_job-821551.txt

[tsalo]

I don't think the problem is that your data have a different image orientation, since XCP-D uses template-space derivatives that should all have the same orientation. The problem is probably because you're using native-resolution derivatives, and your BOLD data probably have slightly different native resolutions that mean that the affines of the template-space derivatives are not the same. Generally, I recommend using preprocessed data with a specific target resolution so that the affines of the template-space outputs are the same across the whole dataset.

[JohannesWiesner]

Hm, could be, thanks! Still, decided to drop this session, because I was not able to swap the images without interpolation. So in a way, it's good, that XCP-D failed here, because otherwise I wouldn't have noticed that this session has images with different orientations.

[JohannesWiesner]

I did a sanity check, all bold-runs have the same native resolution. XCP-D only fails for this one subject and session where all bold runs where acquired with different image orientation.

[tsalo]

Can you share the affines and zooms for the fMRIPrep MNI-space outputs of the two runs with different orientations, as well as one run from another subject that has a consistent orientation?

[JohannesWiesner]

Derived from fmriprep:

sub-XXX_ses-PEZ0X_task-rest_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz

This is the subject-session combo that fails:

PIAMJ5O5J - PEZ02
Affine:

[[   3.    0.    0.  -78.]
 [   0.    5.    0. -117.]
 [   0.    0.    3.  -78.]
 [   0.    0.    0.    1.]] 

Zooms:

(np.float32(3.0), np.float32(5.0), np.float32(3.0), np.float32(2.0)) 

XCP-D runs without errors on the other two sessions from the same subject

PIAMJ5O5J - PEZ01
Affine:

[[   3.    0.    0.  -78.]
 [   0.    3.    0. -114.]
 [   0.    0.    5.  -78.]
 [   0.    0.    0.    1.]] 

Zooms:

(np.float32(3.0), np.float32(3.0), np.float32(5.0), np.float32(2.0)) 

PIAMJ5O5J - PEZ03
Affine:

[[   3.    0.    0.  -78.]
 [   0.    3.    0. -114.]
 [   0.    0.    5.  -78.]
 [   0.    0.    0.    1.]] 

Zooms:

(np.float32(3.0), np.float32(3.0), np.float32(5.0), np.float32(2.0)) 

Here's a different subject where everything also worked fine:

PIAM137OU - PEZ01
Affine:

[[   3.    0.    0.  -78.]
 [   0.    3.    0. -114.]
 [   0.    0.    5.  -78.]
 [   0.    0.    0.    1.]] 

Zooms:

(np.float32(3.0), np.float32(3.0), np.float32(5.0), np.float32(2.0)) 

PIAM137OU - PEZ02
Affine:

[[   3.    0.    0.  -78.]
 [   0.    3.    0. -114.]
 [   0.    0.    5.  -78.]
 [   0.    0.    0.    1.]] 

Zooms:

(np.float32(3.0), np.float32(3.0), np.float32(5.0), np.float32(2.0)) 

[tsalo]

Just to clarify- that's from the native-resolution MNI-space output from fMRIPrep, not native-space outputs?

[JohannesWiesner]

Just to clarify- that's from the native-resolution MNI-space output from fMRIPrep, not native-space outputs?

Sorry, you're right, updated my comment (now derived from fmriprep not rawdata)

[tsalo]

Thanks! The voxel sizes are exactly the same, but somehow the fourth column is different. Can you share the shapes of the images as well?

EDIT: And the shapes/dimensions of the raw files too, if possible.
EDIT 2: For context, I believe this is related to nipreps/fmriprep#3228.