# Improving Interpretability in Medical Imaging Diagnosis using Adversarial Training

arXiv: https://arxiv.org/abs/2012.01166 · HTML: https://mlsanity.com/paper/2012.01166/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 2 Dec 2020.

## Built on by (most cited first, 9 of 9)

- [Interpretable Deep Learning: Interpretation, Interpretability, Trustworthiness, and Beyond](https://mlsanity.com/paper/2103.10689/) (2021)
- [Explainable Deep Learning Methods in Medical Image Classification: A Survey](https://mlsanity.com/paper/2205.04766/) (2022)
- [Label-Assemble: Leveraging Multiple Datasets with Partial Labels](https://mlsanity.com/paper/2109.12265/) (2021)
- [Semi-supervised classification of radiology images with NoTeacher: A Teacher that is not Mean](https://mlsanity.com/paper/2108.04423/) (2021)
- [Improving Interpretability via Regularization of Neural Activation Sensitivity](https://mlsanity.com/paper/2211.08686/) (2022)
- [Enhancing the "Immunity" of Mixture-of-Experts Networks for Adversarial Defense](https://mlsanity.com/paper/2402.18787/) (2024)
- [Robust and Interpretable COVID-19 Diagnosis on Chest X-ray Images using Adversarial Training](https://mlsanity.com/paper/2311.14227/) (2023)
- [Dynamic Perturbation-Adaptive Adversarial Training on Medical Image Classification](https://mlsanity.com/paper/2403.06798/) (2024)
- [Quantitative Evaluation of the Saliency Map for Alzheimer's Disease Classifier with Anatomical Segmentation](https://mlsanity.com/paper/2407.08546/) (2024)

## Abstract (arXiv)

We investigate the influence of adversarial training on the interpretability of convolutional neural networks (CNNs), specifically applied to diagnosing skin cancer. We show that gradient-based saliency maps of adversarially trained CNNs are significantly sharper and more visually coherent than those of standardly trained CNNs. Furthermore, we show that adversarially trained networks highlight regions with significant color variation within the lesion, a common characteristic of melanoma. We find that fine-tuning a robust network with a small learning rate further improves saliency maps' sharpness. Lastly, we provide preliminary work suggesting that robustifying the first layers to extract robust low-level features leads to visually coherent explanations.