Fix: 2D coords into atlas space

[zenWai]

Description

What is this PR

• Bug fix

Why is this PR needed?

Related to 2D:

• Coordinates centered on the slice position instead of atlas space ([BUG] Axes generated by brainglobe-heatmap are potentially wrong #55)
• Coordinates centered on brain center of mass with named orientations
• Coordinates centered on plane center with tuple orientations
• Tuple orientations like (1, 0, 0) not matching named orientation "frontal"
• slicer_2D.py example contours not overlapping with reference ([BUG] slicer_2D.py example is broken #72)
• get_coordinates.py example returning brain center of mass coordinates

What does this PR do?

This PR adds a separate projection step in the Slicer _project_to_2d( ) that:

• Undoes the brainrender Z-flip on intersection points
• Projects to 2D with corrected atlas-space M
• Add example heatmap_2d_subplots_oblique.py

New tests:
test_get_coordinates.py

• Confirms mesh contour coords are within 75µm of atlas annotation
• tuple orientation (1,0,0) produces same coords as "frontal", etc.
• oblique orientations (0,1,1), (1,0,1), (0.3,0.7,1) produce spread on both 2D axes(Confirms that axis don't go flat due to Z)

test_slicer_2d_example.py

• runs the slicer_2D.py example end-to-end, verifies region contours are drawn on top of the brain image

New example:
heatmap_2d_subplots_oblique.py

The get_projections( ) method is preserved for anyone using plane centered coordinates and doc updated.

References

• Fix [BUG] Axes generated by brainglobe-heatmap are potentially wrong #55
• Fix [BUG] slicer_2D.py example is broken #72

How has this PR been tested?

• New tests
• New example
• A custom plan.py example-like to confirm every plane0 projections on 3D match 2D new example images.

Is this a breaking change?

Yes ✔
Expectations of get_structures_slice_coords() changes:

• Code that negated Y coordinates to compensate (like slicer_2D.py did with -cont[:, 1]) will need to remove that negation
• Coordinates are not centered with brain center of mass on named orientations[bug fix]
• Coordinates are not centered with plane center on tuple orientations[bug fix]

Does this PR require an update to the documentation?

No

Checklist:

• The code has been tested locally
• Tests have been added to cover all new functionality (unit & integration)
• The documentation has been updated to reflect any changes
• The code has been formatted with pre-commit

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 77.50%. Comparing base (0fea8e1) to head (a2f1a59).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #110      +/-   ##
==========================================
+ Coverage   69.63%   77.50%   +7.86%     
==========================================
  Files           5        5              
  Lines         326      360      +34     
==========================================
+ Hits          227      279      +52     
+ Misses         99       81      -18     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

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

[zenWai]

@IgorTatarnikov It closes your issues, would be great to get your review and see if the coords are as expected or something still broken. tysm

[IgorTatarnikov]

Hi @zenWai, apologies it took so long for me to look at this.

This looks great! The axis aligned slicing now returns the correct positions.

However, for the oblique case, some of the axes end up displaying as negative. This should not be the case if we're referring to the real world anatomical coordinates. See below for a screenshot.

[IgorTatarnikov]

Is there a way to avoid duplicating this with the function that already exists in plane.py?