Ensembling Neural Networks for Digital Pathology Images Classification and Segmentation

TL;DR

This paper explores ensembling CNNs for digital pathology image classification and segmentation, addressing challenges of large image sizes and limited training data, achieving high accuracy on breast cancer histology images.

Contribution

It introduces a patch-based classification and ensembling approach tailored for digital pathology, improving prediction accuracy despite data limitations.

Findings

Achieved 90% accuracy on breast cancer histology dataset.

Demonstrated effectiveness of patch-based ensembling for large images.

Validated approach on real-world histology data.

Abstract

In the last years, neural networks have proven to be a powerful framework for various image analysis problems. However, some application domains have specific limitations. Notably, digital pathology is an example of such fields due to tremendous image sizes and quite limited number of training examples available. In this paper, we adopt state-of-the-art convolutional neural networks (CNN) architectures for digital pathology images analysis. We propose to classify image patches to increase effective sample size and then to apply an ensembling technique to build prediction for the original images. To validate the developed approaches, we conducted experiments with \textit{Breast Cancer Histology Challenge} dataset and obtained 90\% accuracy for the 4-class tissue classification task.