Structure-Preserving Color Normalization and Sparse Stain Separation for Histological Images

By Abhishek Vahadane, Tingying Peng, A. Sethi, Shadi Albarqouni, Lichao Wang, Maximilian Baust, K. Steiger, A. Schlitter, I. Esposito, Nassir Navab. IEEE Transactions on Medical Imaging, 2016

759 citations recorded by the source corpus.

Staining and scanning of tissue samples for microscopic examination is fraught with undesirable color variations arising from differences in raw materials and manufacturing techniques of stain vendors, staining protocols of labs, and color responses of digital scanners. When comparing tissue samples, color normalization and stain separation of the tissue images can be helpful for both pathologists and software. Techniques that are used for natural images fail to utilize structural properties of stained tissue samples and produce undesirable color distortions. The stain concentration cannot be negative. Tissue samples are stained with only a few stains and most tissue regions are characterized by at most one effective stain. We model these physical phenomena that define the tissue structure by first decomposing images in an unsupervised manner into stain density maps that are sparse and non-negative. For a given image, we combine its stain density maps with stain color basis of a pathologist-preferred target image, thus altering only its color while preserving its structure described by the maps. Stain density correlation with ground truth and preference by pathologists were higher for images normalized using our method when compared to other alternatives. We also propose a computationally faster extension of this technique for large whole-slide images that selects an appropriate patch sample instead of using the entire image to compute the stain color basis.

Subfields: Artificial Intelligence, Biophysics, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/a243c7bbd2affd548e18a35dadf0313c1b2198c2

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Biophysics

In Computer Vision and Pattern Recognition

Citations in the corpus

Pathomic Fusion: An Integrated Framework for Fusing Histopathology and Genomic Features for Cancer Diagnosis and Prognosis (2019, methodology): “The histology region-of-interests from the diagnostic slides were curated from [23] and are of size 1024 × 1024 , and were normalized using sparse stain normalization [53] to match all images to a standard H&E histology image.”
Diffusion models in medical imaging: A comprehensive survey (2022, methodology): “To tackle this data consistency problem and enforce the model to focus more on morphological patterns, they first feed input images into a color normalization module [124] to unify the domain of all images.”
Deep neural network models for computational histopathology: A survey (2019, neutral): “Earlier machine learning (ML) methods (Macenko et al., 2009; Vahadane et al., 2016) assume that staining attenuates light (optical density) uniformly and decompose each optical density image into concentration (appearance) and color (stain) matrices, and use the latter to stain new images by splitti”
Deep Adversarial Training for Multi-Organ Nuclei Segmentation in Histopathology Images (2018, methodology): “Details about this method can be found in [62] and examples of stain normalized images have been shown in”
The impact of pre- and post-image processing techniques on deep learning frameworks: A comprehensive review for digital pathology image analysis (2020, methodology): “Spectral matching [22,58], Non-negative Matrix Factorization (NMF) [59], and Independent Component Analysis (ICA) [60] methods have been employed to estimate both the stain density map H and the stain color appearance matrix W .”
A Multi-Organ Nucleus Segmentation Challenge (2020, methodology): “[35] was used by ten teams due to its demonstrated performance and code avail-”
Measuring Domain Shift for Deep Learning in Histopathology (2020, positive): “Forthisstudy,themethodpresentedin[33]wasused.”
Staingan: Stain Style Transfer for Digital Histological Images (2018, methodology): “To have a fair comprehensive comparison, we evaluate our model against the state of the art methods of Reinhard[11] Macenko[10], Khan[9], and Vahadane[13] as follows; i) Quantitatve comparison between the visual appearances of stained images, and the eﬀect of varying reference slides, ii) Quantitati”