CoRPA: Adversarial Image Generation for Chest X-rays Using Concept Vector Perturbations and Generative Models

TL;DR

This paper introduces CoRPA, a novel adversarial attack method that uses clinical concept perturbations to generate realistic misdiagnosis scenarios in chest X-ray AI models, revealing domain-specific vulnerabilities.

Contribution

The study presents CoRPA, a clinically-focused black-box adversarial attack framework tailored for medical imaging, highlighting vulnerabilities not detected by conventional attacks.

Findings

Deep learning models are less robust to CoRPA than to traditional attacks.

CoRPA generates realistic clinical misdiagnosis scenarios.

Medical AI systems need domain-specific robustness evaluations.

Abstract

Deep learning models for medical image classification tasks are becoming widely implemented in AI-assisted diagnostic tools, aiming to enhance diagnostic accuracy, reduce clinician workloads, and improve patient outcomes. However, their vulnerability to adversarial attacks poses significant risks to patient safety. Current attack methodologies use general techniques such as model querying or pixel value perturbations to generate adversarial examples designed to fool a model. These approaches may not adequately address the unique characteristics of clinical errors stemming from missed or incorrectly identified clinical features. We propose the Concept-based Report Perturbation Attack (CoRPA), a clinically-focused black-box adversarial attack framework tailored to the medical imaging domain. CoRPA leverages clinical concepts to generate adversarial radiological reports and images that closely mirror realistic clinical misdiagnosis scenarios. We demonstrate the utility of CoRPA using the MIMIC-CXR-JPG dataset of chest X-rays and radiological reports. Our evaluation reveals that deep learning models exhibiting strong resilience to conventional adversarial attacks are significantly less robust when subjected to CoRPA's clinically-focused perturbations. This underscores the importance of addressing domain-specific vulnerabilities in medical AI systems. By introducing a specialized adversarial attack framework, this study provides a foundation for developing robust, real-world-ready AI models in healthcare, ensuring their safe and reliable deployment in high-stakes clinical environments.