# Learning Interpretable Features via Adversarially Robust Optimization

## TL;DR

This paper introduces a method to enhance neural network interpretability in medical imaging by integrating adversarial robustness into training, resulting in more understandable features without sacrificing accuracy.

## Contribution

The work presents a novel approach combining adversarially robust optimization with interpretability measures for neural networks in medical diagnosis.

## Key findings

- Improved feature interpretability via weakly supervised localization.
- Enhanced visual interpretability with class activation and saliency maps.
- Maintained state-of-the-art accuracy on chest X-ray classification.

## Abstract

Neural networks are proven to be remarkably successful for classification and diagnosis in medical applications. However, the ambiguity in the decision-making process and the interpretability of the learned features is a matter of concern. In this work, we propose a method for improving the feature interpretability of neural network classifiers. Initially, we propose a baseline convolutional neural network with state of the art performance in terms of accuracy and weakly supervised localization. Subsequently, the loss is modified to integrate robustness to adversarial examples into the training process. In this work, feature interpretability is quantified via evaluating the weakly supervised localization using the ground truth bounding boxes. Interpretability is also visually assessed using class activation maps and saliency maps. The method is applied to NIH ChestX-ray14, the largest publicly available chest x-rays dataset. We demonstrate that the adversarially robust optimization paradigm improves feature interpretability both quantitatively and visually.