Deep Relevance Regularization: Interpretable and Robust Tumor Typing of Imaging Mass Spectrometry Data

TL;DR

This paper introduces Deep Relevance Regularization, a technique to improve neural network robustness and interpretability in tumor typing from imaging mass spectrometry data, especially across different laboratories.

Contribution

The paper proposes Deep Relevance Regularization to restrict neural network focus, enhancing robustness against confounding factors and improving interpretability in multi-lab tumor classification.

Findings

Improves classification accuracy across labs

Produces sparser, more interpretable relevance maps

Enhances robustness to confounding factors

Abstract

Neural networks have recently been established as a viable classification method for imaging mass spectrometry data for tumor typing. For multi-laboratory scenarios however, certain confounding factors may strongly impede their performance. In this work, we introduce Deep Relevance Regularization, a method of restricting what the neural network can focus on during classification, in order to improve the classification performance. We demonstrate how Deep Relevance Regularization robustifies neural networks against confounding factors on a challenging inter-lab dataset consisting of breast and ovarian carcinoma. We further show that this makes the relevance map -- a way of visualizing the discriminative parts of the mass spectrum -- sparser, thereby making the classifier easier to interpret