# Assessing the validity of saliency maps for abnormality localization in medical imaging

arXiv: https://arxiv.org/abs/2006.00063 · HTML: https://mlsanity.com/paper/2006.00063/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 29 May 2020.

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

- [Diffusion Models for Medical Anomaly Detection](https://mlsanity.com/paper/2203.04306/) (2022)
- [Algorithm Fairness in AI for Medicine and Healthcare](https://mlsanity.com/paper/2110.00603/) (2021)
- [Revisiting Sanity Checks for Saliency Maps](https://mlsanity.com/paper/2110.14297/) (2021)
- [Evaluating Feature Attribution Methods in the Image Domain](https://mlsanity.com/paper/2202.12270/) (2022)
- [Gifsplanation via Latent Shift: A Simple Autoencoder Approach to Counterfactual Generation for Chest X-rays](https://mlsanity.com/paper/2102.09475/) (2021)
- [Evaluating the Use of Reconstruction Error for Novelty Localization](https://mlsanity.com/paper/2107.13379/) (2021)
- [Disentangled representations: towards interpretation of sex determination from hip bone](https://mlsanity.com/paper/2112.09414/) (2021)
- [Learning from the few: Fine-grained approach to pediatric wrist pathology recognition on a limited dataset](https://mlsanity.com/paper/2408.13542/) (2024)
- [Introduction to Machine Learning for Physicians: A Survival Guide for Data Deluge](https://mlsanity.com/paper/2212.12303/) (2022)

## Abstract (arXiv)

Saliency maps have become a widely used method to assess which areas of the input image are most pertinent to the prediction of a trained neural network. However, in the context of medical imaging, there is no study to our knowledge that has examined the efficacy of these techniques and quantified them using overlap with ground truth bounding boxes. In this work, we explored the credibility of the various existing saliency map methods on the RSNA Pneumonia dataset. We found that GradCAM was the most sensitive to model parameter and label randomization, and was highly agnostic to model architecture.