Exploring key stakeholders’ perspectives on integrating the EU AI Act with the MDR for certifying AI medical devices

Abstract

AbstractArtificial Intelligence (AI) algorithms are transforming healthcare by advancing the capabilities of medical devices. These algorithms can now analyze X-ray images to help detect and interpret medical conditions, particularly in radiology and pathology. The growing use of AI in medicine highlights the importance of strict regulations to prioritize patient safety during the development and deployment of AI in medical devices. Despite extensive research, there are limited empirical studies on stakeholders’ perspectives regarding the challenges of implementing AI regulatory frameworks alongside existing legally binding regulations specific to medical devices. This study uses semi-structured interviews to explore the perspectives of key stakeholders, such as regulatory bodies and medical device manufacturers, to understand the potential implications of integrating the European Union Artificial Intelligence Act with the Medical Device Regulation to certify AI-embedded medical devices. Through inductive-thematic analysis and adopting activity theory to further synthesize the interviews, the findings reveal stakeholders’ challenges, including uncertainty about implementing different regulations, resource limitations, and the potential impact of complex regulations on healthcare quality. Additionally, the results indicate that key stakeholders did not explicitly identify the absence of a comprehensive ethical framework for AI-enabled medical devices in both regulations as a major challenge. This underscores the need to develop a harmonized ethical framework for certifying AI used in healthcare.