MedAIScout: Automated Retrieval of Known Machine Learning Vulnerabilities in Medical Applications

Athish Pranav Dharmalingam, Gargi Mitra

Keywords: AI red-teaming, medical device security, ML-enabled medical device, automated information retrieval, ML attacks

TL;DR: This paper proposes MedAIScout, a semi-automated tool that streamlines AI red teaming efforts by quickly retrieving critical information on ML vulnerabilities specific to a given ML-enabled medical device.

Abstract: Machine learning (ML)-enabled medical devices are transforming the healthcare industry but are vulnerable to adversarial attacks that can compromise their safety. Current red teaming efforts often overlook these ML-specific threats, leaving devices exposed. To address this, we present MedAIScout, a semi-automated tool designed to retrieve information on known ML vulnerabilities relevant to ML-enabled medical devices. Through case studies on five FDA-approved ML-enabled devices, we demonstrate that MedAIScout effectively identifies relevant vulnerabilities, significantly aiding red teaming efforts

Serve As Reviewer: No preferred reviewer

Submission Number: 53