A method for normalizing histology slides for quantitative analysis

By Marc Macenko, Marc Niethammer, James Stephen Marron, D. Borland, J. Woosley, Xiaojun Guan, Charles Schmitt, N. Thomas. IEEE International Symposium on Biomedical Imaging, 2009

1,467 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Biophysics, Molecular Biology

View Paper: https://www.semanticscholar.org/paper/ed81378eb0937a6425f6221800049c03dd8dd114

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 Molecular Biology

Citations in the corpus

BACH: Grand challenge on breast cancer histology images (2018, neutral): “77 ResNet-50, ResNet-101 3 2 7 1 224×224 [30]”
Structure-Preserving Color Normalization and Sparse Stain Separation for Histological Images (2016, methodology): “In this paper, we have compared the proposed SNMF-based stain separation with NMF-based [12] and SVD-based [9] methods.”
The impact of pre- and post-image processing techniques on deep learning frameworks: A comprehensive review for digital pathology image analysis (2020, methodology): “The automated extraction of the matrix W has been done by using: (i) prior information of the stain vectors [52], (ii) singular value decomposition [51,53] (iii) support vector machines [25], (iv) gaussian mixture model [54] and (v) cellular structures segmentation [55 – 57].”
Deep Adversarial Training for Multi-Organ Nuclei Segmentation in Histopathology Images (2018, methodology): “We empirically found that sparse stain normalized H&E images following [60] performed better with our nuclei segmentation methods as compared to standard Reinhard [56] and Macenko [59] stain normalization.”
A deep learning algorithm for one-step contour aware nuclei segmentation of histopathology images (2018, background): “A few H&E stain normalization methods [8, 16, 31] have been proposed to eliminate the negative interference caused by color variation.”
Deep Convolutional Neural Networks for Breast Cancer Histology Image Analysis (2018, methodology): “To bring the microscopy images into a common space to enable improved quantitative analysis, we normalize the amount of H&E stained on the tissue as described in [22].”
Methods for Nuclei Detection, Segmentation, and Classification in Digital Histopathology: A Review—Current Status and Future Potential (2014, background): “For instance, the overall best results on mitosis detection presented during the recent MITOS and AMIDA contests achieved an F-score of 78.21% for MITOS [71] and 61.1% for AMIDA [112], which would not be considered accurate detection under medical terms.”
Staingan: Stain Style Transfer for Digital Histological Images (2018, methodology): “Our proposed StainGAN method shows a relative improvement of ( 80% , 36% , 22% , 15% , and 3% ) on AUC over the Un-normalized, Reinhard, Macenko, Khan, and Vahadane, respectively.”
Deep neural network models for computational histopathology: A survey (2019, positive): “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 Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support (2017, positive): “For hematoxylin and eosin (H&E) stained slides, staining normalization (SN) methods can be used [7,6].”
Predicting survival from colorectal cancer histology slides using deep learning: A retrospective multicenter study (2019, positive): “All imagewere Macenko-normalizedbeforefurther analysis[34].”
AggNet: Deep Learning From Crowds for Mitosis Detection in Breast Cancer Histology Images (2016, positive): “pre-processed by staining appearance normalization [33].”
Harnessing multimodal data integration to advance precision oncology (2021, methodology): “Preprocessing techniques in MRI 122 and H&E staining 123,124 address this heterogeneity, and with large cohorts, DL is somewhat robust with regard to noise 28,125 , but such heterogeneity is a major reason that AI systems fail when trialled in the clinic 126 .”