Learning to Segment Breast Biopsy Whole Slide Images

TL;DR

This paper introduces a novel encoder-decoder model with four key modifications for effective semantic segmentation of large breast biopsy images, improving accuracy over traditional methods and enhancing automated diagnosis.

Contribution

The paper presents four innovative modifications to encoder-decoder networks enabling segmentation of large biopsy images, addressing challenges of structure size variation and image scale.

Findings

Model outperforms feature-based and conventional encoder-decoder approaches

Semantic segmentation improves diagnostic accuracy

Multi-resolution fusion enhances segmentation quality

Abstract

We trained and applied an encoder-decoder model to semantically segment breast biopsy images into biologically meaningful tissue labels. Since conventional encoder-decoder networks cannot be applied directly on large biopsy images and the different sized structures in biopsies present novel challenges, we propose four modifications: (1) an input-aware encoding block to compensate for information loss, (2) a new dense connection pattern between encoder and decoder, (3) dense and sparse decoders to combine multi-level features, (4) a multi-resolution network that fuses the results of encoder-decoders run on different resolutions. Our model outperforms a feature-based approach and conventional encoder-decoders from the literature. We use semantic segmentations produced with our model in an automated diagnosis task and obtain higher accuracies than a baseline approach that employs an SVM for feature-based segmentation, both using the same segmentation-based diagnostic features.