VahadaneExtractor

class VahadaneExtractor(luminosity_threshold=0.8, regularizer=0.1)[source]

Vahadane stain extractor.

Get the stain matrix as defined in:

Vahadane, Abhishek, et al. “Structure-preserving color normalization and sparse stain separation for histological images.” IEEE transactions on medical imaging 35.8 (2016): 1962-1971.

This class contains code inspired by StainTools [https://github.com/Peter554/StainTools] written by Peter Byfield.

Warning

Vahadane stain extraction/normalization algorithms are unstable after the update to dictionary learning algorithm in scikit-learn > v0.23.0 (see issue #382). Please be advised and consider using other stain extraction (normalization) algorithms or toolboxes, such as https://github.com/CielAl/torch-staintools

Parameters:

  • luminosity_threshold (float) – Threshold used for tissue area selection.

  • regularizer (float) – Regularizer used in dictionary learning.

Examples

>>> from tiatoolbox.tools.stainextract import VahadaneExtractor
>>> from tiatoolbox.utils import imread
>>> extractor = VahadaneExtractor()
>>> img = imread('path/to/image')
>>> stain_matrix = extractor.get_stain_matrix(img)

Initialize VahadaneExtractor.

Methods

get_stain_matrix

Stain matrix estimation.
get_stain_matrix(img)[source]

Stain matrix estimation.

Parameters:
Returns:

Estimated stain matrix.

Return type:

numpy.ndarray