# Hidden in Plain Sight: Undetectable Adversarial Bias Attacks on Vulnerable Patient Populations

arXiv: https://arxiv.org/abs/2402.05713 · HTML: https://mlsanity.com/paper/2402.05713/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 6 of its 25 references that are in the map.
- First posted on arXiv 8 Feb 2024.

## Builds on (most cited first, 6 of 6)

- [Identifying and Correcting Label Bias in Machine Learning](https://mlsanity.com/paper/1901.04966/) (2019)
- [Adversarial Attacks Against Medical Deep Learning Systems](https://mlsanity.com/paper/1804.05296/) (2018)
- [On the limits of cross-domain generalization in automated X-ray prediction](https://mlsanity.com/paper/2002.02497/) (2020)
- [Improving the Fairness of Chest X-ray Classifiers](https://mlsanity.com/paper/2203.12609/) (2022)
- [Label Poisoning is All You Need](https://mlsanity.com/paper/2310.18933/) (2023)
- [A Strong Baseline for Domain Adaptation and Generalization in Medical Imaging](https://mlsanity.com/paper/1904.01638/) (2019)

## Abstract (arXiv)

The proliferation of artificial intelligence (AI) in radiology has shed light on the risk of deep learning (DL) models exacerbating clinical biases towards vulnerable patient populations. While prior literature has focused on quantifying biases exhibited by trained DL models, demographically targeted adversarial bias attacks on DL models and its implication in the clinical environment remains an underexplored field of research in medical imaging. In this work, we demonstrate that demographically targeted label poisoning attacks can introduce undetectable underdiagnosis bias in DL models. Our results across multiple performance metrics and demographic groups like sex, age, and their intersectional subgroups show that adversarial bias attacks demonstrate high-selectivity for bias in the targeted group by degrading group model performance without impacting overall model performance. Furthermore, our results indicate that adversarial bias attacks result in biased DL models that propagate prediction bias even when evaluated with external datasets.