mlsanityData to 7 Oct 2026

Understanding Adversarial Attacks on Deep Learning Based Medical Image Analysis Systems

arXiv 1907.10456 ↗First posted 24 Jul 2019cs.CV cs.LG eess.IVcited by 151

Not on the map. Why?

Abstract (arXiv)

Deep neural networks (DNNs) have become popular for medical image analysis tasks like cancer diagnosis and lesion detection. However, a recent study demonstrates that medical deep learning systems can be compromised by carefully-engineered adversarial examples/attacks with small imperceptible perturbations. This raises safety concerns about the deployment of these systems in clinical settings. In this paper, we provide a deeper understanding of adversarial examples in the context of medical images. We find that medical DNN models can be more vulnerable to adversarial attacks compared to models for natural images, according to two different viewpoints. Surprisingly, we also find that medical adversarial attacks can be easily detected, i.e., simple detectors can achieve over 98% detection AUC against state-of-the-art attacks, due to fundamental feature differences compared to normal examples. We believe these findings may be a useful basis to approach the design of more explainable and secure medical deep learning systems.

Key facts. Cited by 151 papers in the map (data through 7 Oct 2026). First posted on arXiv 24 Jul 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 34; 2022: 37; 2023: 22; 2024: 26; 2025: 26; 2026: 6.

151 papers.

Sort:

By area:

  1. 2021Neural Attention Distillation: Erasing Backdoor Triggers from Deep Neural Networkscited by 313
  2. 2022Transformers in Medical Imaging: A Surveycited by 245
  3. 2021VT-ADL: A Vision Transformer Network for Image Anomaly Detection and Localizationcited by 222
  4. 2021Unlearnable Examples: Making Personal Data Unexploitablecited by 188
  5. 2023Sociotechnical Safety Evaluation of Generative AI Systemscited by 154
  6. 2021Revisiting Adversarial Robustness Distillation: Robust Soft Labels Make Student Bettercited by 77
  7. 2021Improving Adversarial Robustness via Channel-wise Activation Suppressingcited by 76
  8. 2021When Does Contrastive Learning Preserve Adversarial Robustness from Pretraining to Finetuning?cited by 76
  9. 2021Exploring Architectural Ingredients of Adversarially Robust Deep Neural Networkscited by 73
  10. 2022Trustworthy clinical AI solutions: a unified review of uncertainty quantification in deep learning models for medical image analysiscited by 62
  11. 2023CFA: Class-wise Calibrated Fair Adversarial Trainingcited by 53
  12. 2024Explainable Artificial Intelligence: A Survey of Needs, Techniques, Applications, and Future Directioncited by 46
38 more
  1. 2022Self-Ensemble Adversarial Training for Improved Robustnesscited by 38
  2. 2022DCSAU-Net: A Deeper and More Compact Split-Attention U-Net for Medical Image Segmentationcited by 37
  3. 2022Towards Understanding and Boosting Adversarial Transferability from a Distribution Perspectivecited by 36
  4. 2022SAFE: Sensitivity-Aware Features for Out-of-Distribution Object Detectioncited by 33
  5. 2022The Many Faces of Adversarial Riskcited by 28
  6. 2022A Survey of Adversarial Defences and Robustness in NLPcited by 22
  7. 2021Removing Adversarial Noise in Class Activation Feature Spacecited by 18
  8. 2023Survey on Adversarial Attack and Defense for Medical Image Analysis: Methods and Challengescited by 16
  9. 2023Adversarial Examples Are Not Real Featurescited by 16
  10. 2021Towards Evaluating the Robustness of Deep Diagnostic Models by Adversarial Attackcited by 15
  11. 2021Explainable Deep Learning in Healthcare: A Methodological Survey from an Attribution Viewcited by 15
  12. 2021Towards Robust General Medical Image Segmentationcited by 14
  13. 2023Sharpness-Aware Minimization Alone can Improve Adversarial Robustnesscited by 14
  14. 2022Improving Adversarial Robustness via Mutual Information Estimationcited by 13
  15. 2024The Effect of Intrinsic Dataset Properties on Generalization: Unraveling Learning Differences Between Natural and Medical Imagescited by 13
  16. 2021Fooling Adversarial Training with Inducing Noisecited by 12
  17. 2021ℓ_∞-Robustness and Beyond: Unleashing Efficient Adversarial Trainingcited by 11
  18. 2022Improving Adversarial Robustness by Putting More Regularizations on Less Robust Samplescited by 11
  19. 2021Stabilized Medical Image Attackscited by 10
  20. 2021When and How to Fool Explainable Models (and Humans) with Adversarial Examplescited by 10
  21. 2024Game-Theoretic Defenses for Robust Conformal Prediction Against Adversarial Attacks in Medical Imagingcited by 10
  22. 2021MixACM: Mixup-Based Robustness Transfer via Distillation of Activated Channel Mapscited by 9
  23. 2021MedRDF: A Robust and Retrain-Less Diagnostic Framework for Medical Pretrained Models Against Adversarial Attackcited by 9
  24. 2022Adversarial Robustness through the Lens of Convolutional Filterscited by 9
  25. 2022Adversarial Coreset Selection for Efficient Robust Trainingcited by 9
  26. 2023It Is All About Data: A Survey on the Effects of Data on Adversarial Robustnesscited by 9
  27. 2021What Do Deep Nets Learn? Class-wise Patterns Revealed in the Input Spacecited by 8
  28. 2021Modeling Adversarial Noise for Adversarial Trainingcited by 8
  29. 2022VisCUIT: Visual Auditor for Bias in CNN Image Classifiercited by 8
  30. 2022Deep Learning for Medical Image Registration: A Comprehensive Reviewcited by 8
  31. 2022Explainable, Domain-Adaptive, and Federated Artificial Intelligence in Medicinecited by 8
  32. 2023Exploring adversarial attacks in federated learning for medical imagingcited by 8
  33. 2024Enhancing Adversarial Robustness of Vision-Language Models through Low-Rank Adaptationcited by 8
  34. 2021Adversarial Attacks and Defenses in Physiological Computing: A Systematic Reviewcited by 7
  35. 2021Attack-agnostic Adversarial Detection on Medical Data Using Explainable Machine Learningcited by 7
  36. 2021Now You See It, Now You Dont: Adversarial Vulnerabilities in Computational Pathologycited by 7
  37. 2022Understanding the Vulnerability of Skeleton-based Human Activity Recognition via Black-box Attackcited by 7
  38. 2024AdvQDet: Detecting Query-Based Adversarial Attacks with Adversarial Contrastive Prompt Tuningcited by 7

50 most cited of 151 All 151 (JSON)