🤖 AI Summary
This work addresses critical gaps in medical AI security and privacy research—namely, uneven coverage of clinical scenarios, threat models misaligned with real-world healthcare practice, and insufficient engagement with the biomedical community. To bridge these gaps, we propose the first cross-domain security and privacy (S&P) risk taxonomy for medical AI applications. Our methodology integrates a systematic literature review, threat modeling grounded in clinical imaging and electronic health records, and feasibility validation via adversarial attacks. We identify core attack surfaces and defense gaps across six major AI-driven healthcare subdomains. The study exposes a structural disconnect between existing threat modeling approaches and actual clinical workflows. Furthermore, we introduce a reusable risk analysis framework and an evaluation benchmark, providing both theoretical foundations and practical guidance for future medical AI security research and development.
📝 Abstract
The integration of artificial intelligence (AI) and machine learning (ML) into healthcare systems holds great promise for enhancing patient care and care delivery efficiency; however, it also exposes sensitive data and system integrity to potential cyberattacks. Current security and privacy (S&P) research on healthcare AI is highly unbalanced in terms of healthcare deployment scenarios and threat models, and has a disconnected focus with the biomedical research community. This hinders a comprehensive understanding of the risks that healthcare AI entails. To address this gap, this paper takes a thorough examination of existing healthcare AI S&P research, providing a unified framework that allows the identification of under-explored areas. Our survey presents a systematic overview of healthcare AI attacks and defenses, and points out challenges and research opportunities for each AI-driven healthcare application domain. Through our experimental analysis of different threat models and feasibility studies on under-explored adversarial attacks, we provide compelling insights into the pressing need for cybersecurity research in the rapidly evolving field of healthcare AI.