Add stain augment

[ternaus]

Summary by Sourcery

Add an H&E stain augmentation to simulate different staining conditions in histopathology images. The augmentation supports predefined stain matrices, Vahadane and Macenko stain extraction methods, and custom matrices. Control the intensity variations with intensity_scale_range and intensity_shift_range, and choose whether to augment background regions with augment_background.

New Features:

• Added HEStain augmentation for histopathology images.

Tests:

• Added tests for the new HEStain augmentation, including shape validation, comparison with scikit-learn NMF, stain separation accuracy, and background augmentation behavior.

[sourcery-ai]

Reviewer's Guide by Sourcery

This PR introduces a new stain augmentation feature for histopathology images. The changes involve adding functions and classes for extracting and normalizing H&E stains using various methods (Vahadane, Macenko, presets), integrating a new HEStain transform into the augmentation pipeline, and expanding the test suite to validate the new functionality.

Updated class diagram for H&E Stain Augmentation

classDiagram
    class StainNormalizer {
        +stain_matrix_target
        +fit(img: np.ndarray)
    }

    class VahadaneNormalizer {
        +fit(img: np.ndarray)
    }

    class MacenkoNormalizer {
        -angular_percentile: float
        +fit(img: np.ndarray, angular_percentile: float = 99)
    }

    StainNormalizer <|-- VahadaneNormalizer
    StainNormalizer <|-- MacenkoNormalizer

    class SimpleNMF {
        -n_iter: int
        -initial_colors: np.array
        +fit_transform(optical_density: np.ndarray) : tuple[np.ndarray, np.ndarray]
    }

    class HEStain {
        -method: Literal["preset", "random_preset", "vahadane", "macenko"]
        -preset: Literal["ruifrok", "macenko", "standard", "high_contrast", "h_heavy", "e_heavy", "dark", "light"]
        -intensity_scale_range: tuple[float, float]
        -intensity_shift_range: tuple[float, float]
        -augment_background: bool
        -stain_normalizer
        -preset_names: list
        +get_stain_matrix(img: np.ndarray) : np.ndarray
        +apply(img: np.ndarray, stain_matrix: np.ndarray, scale_factors: np.ndarray, shift_values: np.ndarray) : np.ndarray
        +get_params_dependent_on_data(params: dict, data: dict) : dict
        +get_transform_init_args() : dict
    }

    class InitSchema {
        -method
        -preset
        -intensity_scale_range: tuple[float, float]
        -intensity_shift_range: tuple[float, float]
        -augment_background: bool
        +validate_matrix_selection() : Self
    }

    HEStain "1" *-- "1" InitSchema : uses

    %% Depending on the method, HEStain composes a stain extractor
    HEStain o-- VahadaneNormalizer : "stain_extractor (if method=='vahadane')"
    HEStain o-- MacenkoNormalizer : "stain_extractor (if method=='macenko')"

Loading

File-Level Changes

Change	Details	Files
Implemented core stain augmentation functions and classes	Added standard reference H&E stain matrices Introduced get_normalizer function to select between Vahadane and Macenko methods Defined base class StainNormalizer along with VahadaneNormalizer and MacenkoNormalizer implementations Implemented SimpleNMF class for non-negative matrix factorization with iterative updates Created order_stains_combined function for robust stain ordering using angular and spectral information Implemented get_tissue_mask helper to identify tissue regions Developed apply_he_stain_augmentation to adjust stain concentrations and reconstruct corrected images	albumentations/augmentations/functional.py
Expanded HE stain augmentation test coverage	Added tests for SimpleNMF, comparing shape, non-negativity, and unit norm constraints Included tests to validate consistency against scikit-learn's NMF implementation Created tests for normalizer output shape and stain separation using synthetic H&E images Added tests for varying angular percentiles in MacenkoNormalizer Tested augmentation effects on stain intensities with both background augmentation and non-background augmentation scenarios	tests/functional/test_functional.py
Integrated new HEStain transform into augmentation pipeline	Added HEStain class in the transforms module to wrap the stain augmentation logic Implemented transform initialization schema with validation for method parameters and preset selection Implemented get_stain_matrix to decide stain matrix selection (preset, random, or extraction via normalizers) Integrated apply method to call apply_he_stain_augmentation and batch transformation methods for images and volumes Modified get_params_dependent_on_data to generate random scale and shift factors	albumentations/augmentations/transforms.py
Updated augmentation test configurations to include the new HEStain transform	Added HEStain to lists of transforms in test augmentation files Included HEStain configurations in augmentation definitions	tests/test_augmentations.py tests/aug_definitions.py tests/test_core.py

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!
• Generate a plan of action for an issue: Comment @sourcery-ai plan on
  an issue to generate a plan of action for it.

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.

[sourcery-ai]

Hey @ternaus - I've reviewed your changes - here's some feedback:

Overall Comments:

• Consider adding a usage example to the documentation for the new transform.
• The SimpleNMF class could benefit from more detailed comments explaining the iterative update steps.

Here's what I looked at during the review

• 🟡 General issues: 1 issue found
• 🟢 Security: all looks good
• 🟡 Testing: 1 issue found
• 🟡 Complexity: 1 issue found
• 🟢 Documentation: all looks good

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}