The Hidden Adversarial Vulnerabilities of Medical Federated Learning
Not on the map. Why?
Abstract (arXiv)
In this paper, we delve into the susceptibility of federated medical image analysis systems to adversarial attacks. Our analysis uncovers a novel exploitation avenue: using gradient information from prior global model updates, adversaries can enhance the efficiency and transferability of their attacks. Specifically, we demonstrate that single-step attacks (e.g. FGSM), when aptly initialized, can outperform the efficiency of their iterative counterparts but with reduced computational demand. Our findings underscore the need to revisit our understanding of AI security in federated healthcare settings.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 11 of its 16 references that are in the map. First posted on arXiv 21 Oct 2023.
Builds on
11 of 16 references.
References · 11
- 2016Practical Black-Box Attacks against Machine Learningcited by 903
- 2019On Evaluating Adversarial Robustnesscited by 447
- 2018Rotation Equivariant CNNs for Digital Pathologycited by 422
- 2019You Only Propagate Once: Accelerating Adversarial Training via Maximal Principlecited by 183
- 2019Understanding Adversarial Attacks on Deep Learning Based Medical Image Analysis Systemscited by 153
- 20158-Bit Approximations for Parallelism in Deep Learningcited by 56
- 2019Efficient Adversarial Training with Transferable Adversarial Examplescited by 56
- 2020Adversarial Attack Vulnerability of Medical Image Analysis Systems: Unexplored Factorscited by 40
- 2023Exploring adversarial attacks in federated learning for medical imagingcited by 8
- 2023A Comparative Study of Federated Learning Models for COVID-19 Detectioncited by 3
- 2022AI Technical Considerations: Data Storage, Cloud usage and AI Pipelinecited by 2
Built on by
Citing papers per year: 2024: 1.
1 paper.