Semi-Supervised Contrastive VAE for Disentanglement of Digital Pathology Images

TL;DR

This paper introduces a novel semi-supervised contrastive variational autoencoder designed for disentangling complex pathology images, specifically enhancing interpretability and generalization in tumor-infiltrating lymphocytes detection.

Contribution

It is the first to apply disentanglement techniques to pathology images, proposing a new architecture with cascading disentanglement and reconstruction branches for improved interpretability.

Findings

Achieved superior performance on complex pathology images.

Enhanced interpretability of deep learning models for TIL detection.

Improved generalization power of the models.

Abstract

Despite the strong prediction power of deep learning models, their interpretability remains an important concern. Disentanglement models increase interpretability by decomposing the latent space into interpretable subspaces. In this paper, we propose the first disentanglement method for pathology images. We focus on the task of detecting tumor-infiltrating lymphocytes (TIL). We propose different ideas including cascading disentanglement, novel architecture, and reconstruction branches. We achieve superior performance on complex pathology images, thus improving the interpretability and even generalization power of TIL detection deep learning models. Our codes are available at https://github.com/Shauqi/SS-cVAE.