Interpretation of Disease Evidence for Medical Images Using Adversarial Deformation Fields

TL;DR

This paper introduces DeFI-GAN, a novel adversarial method that generates spatial explanations of disease evidence in medical images by creating deformation fields that transform diseased images into healthy ones, aiding interpretability.

Contribution

The paper presents DeFI-GAN, a new approach for visualizing disease evidence through deformation fields, improving interpretability over existing coarse or noisy methods.

Findings

DeFI-GAN effectively highlights disease biomarkers in chest X-rays and brain MRIs.

The method outperforms baseline difference maps in longitudinal disease evidence analysis.

It uncovers potential dataset biases and novel disease indicators.

Abstract

The high complexity of deep learning models is associated with the difficulty of explaining what evidence they recognize as correlating with specific disease labels. This information is critical for building trust in models and finding their biases. Until now, automated deep learning visualization solutions have identified regions of images used by classifiers, but these solutions are too coarse, too noisy, or have a limited representation of the way images can change. We propose a novel method for formulating and presenting spatial explanations of disease evidence, called deformation field interpretation with generative adversarial networks (DeFI-GAN). An adversarially trained generator produces deformation fields that modify images of diseased patients to resemble images of healthy patients. We validate the method studying chronic obstructive pulmonary disease (COPD) evidence in chest x-rays (CXRs) and Alzheimer's disease (AD) evidence in brain MRIs. When extracting disease evidence in longitudinal data, we show compelling results against a baseline producing difference maps. DeFI-GAN also highlights disease biomarkers not found by previous methods and potential biases that may help in investigations of the dataset and of the adopted learning methods.