H&E-adversarial network: a convolutional neural network to learn stain-invariant features through Hematoxylin & Eosin regression

TL;DR

This paper introduces H&E-adversarial CNN, a novel deep learning approach that learns stain-invariant features to improve the generalization of histopathology image classification across diverse stain variations.

Contribution

The paper presents a new CNN training method leveraging H&E matrix information to enhance stain-invariance, addressing a key challenge in computational pathology.

Findings

H&E-adversarial CNN outperforms five existing methods in stain heterogeneity tasks.

The approach improves classification accuracy across eleven diverse datasets.

It demonstrates better generalization on multi-center histopathology images.

Abstract

Computational pathology is a domain that aims to develop algorithms to automatically analyze large digitized histopathology images, called whole slide images (WSI). WSIs are produced scanning thin tissue samples that are stained to make specific structures visible. They show stain colour heterogeneity due to different preparation and scanning settings applied across medical centers. Stain colour heterogeneity is a problem to train convolutional neural networks (CNN), the state-of-the-art algorithms for most computational pathology tasks, since CNNs usually underperform when tested on images including different stain variations than those within data used to train the CNN. Despite several methods that were developed, stain colour heterogeneity is still an unsolved challenge that limits the development of CNNs that can generalize on data from several medical centers. This paper aims to present a novel method to train CNNs that better generalize on data including several colour variations. The method, called H&E-adversarial CNN, exploits H&E matrix information to learn stain-invariant features during the training. The method is evaluated on the classification of colon and prostate histopathology images, involving eleven heterogeneous datasets, and compared with five other techniques used to handle stain colour heterogeneity. H&E-adversarial CNNs show an improvement in performance compared to the other algorithms, demonstrating that it can help to better deal with stain colour heterogeneous images.