Transfert functions for reorient

[frheault]

Quick description

New functions to handle reorientations of nibabel images to get Numpy array where L is L, R is R, etc.
...

Type of change

Check the relevant options.

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

Provide data, screenshots, command line to test (if relevant)

...

Checklist

• My code follows the style guidelines of this project (run autopep8)
• I added relevant citations to scripts, modules and functions docstrings and descriptions
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I moved all functions from the script file (except the argparser and main) to scilpy modules
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes

[frheault]

This is an example of what could happen if someone is mixing stride on normal volume (no fodf, no peaks, just voxel) and then try to visualize it. If you use a 1,2,3 image and a 1,2,3 wmparc all is good all the time. But if you use an -1,-2,3 image and a 1,2,3 wmparc you are at risk unless you use the new function.

[frheault]

This is now aim to merge into master.

Introducing the SFI : StatefulImage
You can look at an example here!

(The SFI default to RAS)

I highly suggest someone tries it with my example. I will modify a single script (scil_volume_crop maybe?) later.

[codecov]

Codecov Report

❌ Patch coverage is 97.59036% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 72.60%. Comparing base (0ba0902) to head (a6c4fae).

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1295      +/-   ##
==========================================
+ Coverage   72.51%   72.60%   +0.08%     
==========================================
  Files         295      297       +2     
  Lines       25442    25522      +80     
  Branches     3565     3577      +12     
==========================================
+ Hits        18449    18529      +80     
+ Misses       5488     5487       -1     
- Partials     1505     1506       +1     

Flag	Coverage Δ
smoketests	69.49% <1.20%> (-0.22%)	⬇️
unittests	13.50% <96.38%> (+0.27%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Components	Coverage Δ
Scripts	75.38% <ø> (ø)
Library	69.29% <97.59%> (+0.21%)	⬆️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[EmmaRenauld]

Seems great! I did not test, but it would be easier with a script. Maybe we can merge and test more when the first script is created?

[Copilot]

Pull request overview

This PR introduces new functionality for handling reorientations of nibabel images with state management to preserve original orientation information. The implementation allows loading images in a specific orientation (e.g., RAS) for processing while maintaining the ability to save them back in their original orientation.

Changes:

• Added StatefulImage class that extends nib.Nifti1Image to track original and current orientations
• Added validate_axcodes utility function to validate orientation axis codes
• Added comprehensive test suites for both new modules
• Minor code style improvements (encoding parameters, quote consistency, whitespace)

Reviewed changes

Copilot reviewed 7 out of 8 changed files in this pull request and generated 9 comments.

Show a summary per file

File	Description
src/scilpy/io/stateful_image.py	New class for managing image orientation state with load/save/reorient methods
src/scilpy/utils/orientation.py	New utility function for validating axis codes (L/R, A/P, S/I)
src/scilpy/io/tests/test_stateful_image.py	Comprehensive test suite for StatefulImage class functionality
src/scilpy/utils/tests/test_orientation.py	Test suite for validate_axcodes function
src/scilpy/utils/scilpy_bot.py	Added explicit UTF-8 encoding to subprocess and file operations
src/scilpy/utils/init.py	Changed quote style from single to double quotes for consistency
src/scilpy/io/image.py	Removed extra blank line for code cleanup

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[CHrlS98]

I like it. Here is my first round of comments. An example will help understanding the code. Also, are the peaks extracted using scilpy displayed correctly inside mi-brain even when the stride is different from [1,2,3,4] ? If so will the new method break the interoperability with mi-brain?

[frheault]

@CHrlS98 @EmmaRenauld @arnaudbore This can be tested with various stride (flip axis LPS/RAS, but also swap axis PSL, SAR) for the scripts (scil_volume_reshape, scil_volume_crop, scil_volume_resample, scil_volume_reslice_to_reference)

Try random transformations in different axis and compared to the same command on master, and compare results to see which is intuitive/right.

I think it is important to try really random operation that we would never try (change stride to -2,1,3, resample to 2mmx0.5mmx1mm and see if this results in what we want on master vs this branch.

Use the link above for a ZIP, try the same operation on a image with a crazy stride and a wmparc with a normal stride and check if they end up correct. If you 3 could try something weird (without coordination) and describe what you did, what it gave, what it should have been, I think we will make this work !

[EmmaRenauld]

Here is a first test I did: Using a T1 with stride -1, 2, 3 in mrinfo, appears as LAS with scil_header_print_info. Flipping in y:

• scil_volume_flip: flipped in mi-brain, and mrinfo shows the same info as the original T1. Still appears as LAS.
• mrtransform -flip: flipped in mi-brain, but mrinfo a new stride, -1, -2, 3. Appears as LPS in our header_print. The affine as an extra -1 in y.

It that against our goal? We don't seem to have the same definition of "flipping"

Will make other tests.

[EmmaRenauld]

to_orientation='RAS' is your default value, so I would not add it (here and in other calls below)

[EmmaRenauld]

When I wanted to add type hints with this format, Arnaud told me that it was not our style in scilpy, we only put it in the docstring below. Are we reviewing our preferences? I personnally love it (we do it a lot in dwi_ml) but it's true that we don't use it elsewhere.

[frheault]

Here is a first test I did: Using a T1 with stride -1, 2, 3 in mrinfo, appears as LAS with scil_header_print_info. Flipping in y:

• scil_volume_flip: flipped in mi-brain, and mrinfo shows the same info as the original T1. Still appears as LAS.
• mrtransform -flip: flipped in mi-brain, but mrinfo a new stride, -1, -2, 3. Appears as LPS in our header_print. The affine as an extra -1 in y.

It that against our goal? We don't seem to have the same definition of "flipping"

Will make other tests.

Ahh excellent question, do not use scil_volume_flip for this testing, the scilpy script flip the data and leave the header identical (flip the numpy array). I did not explore the scil_volume_flip script yet, this will be another round of changes.

[frheault]

@arnaudbore I changed the base so it points to dev3.0.x, could you update dev3.0.x with master to avoid conflict / large diff?
(non-urgent)