ANALYZING ADVERSE EVENTS OF FDA APPROVED AI-ENABLED MEDICAL DEVICES FOR RISK-AWARE DEVELOPMENT

Date

2025-11-26

Author

Yang, Annie

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

190
views

0
downloads

Cite This

There is increased research interest and development of artificial intelligence (AI) in various industries, especially in healthcare. Regulatory bodies are evolving to promote patient safety and product efficacy. As theories and regulations advance to guide AI development, there is a need to continuously improve by understanding the adverse events of AI-enabled medical devices. Adverse events can reflect inadequacies in the AI system life cycle processes. AI-enabled medical devices life cycle processes must be further examined after deployment and in their operational environments to understand their complex nature and the compounding impact of unexamined deviations from processes which can impact the overall system and product. A comprehensive life cycle process approach is crucial to properly evaluate and analyze where and how failures, errors, and risks arise in AI-enabled medical devices to improve risk mitigation strategies. This study aimed to guide AI medical device stakeholders to be aware of risks in the development process. To achieve this objective, the study investigated the approved 950 AI-enabled medical devices by the United States Food and Drug Administration (US FDA) as of 2024. Using content analysis, the AI-related adverse events of the FDA medical devices were mapped to the AI life cycle processes as detailed in the ISO/IEC 5338: Artificial intelligence — AI System Life Cycle Processes. A set of questions was created based on the represented life cycle processes for stakeholders as a risk mitigation strategy. The question framework aims to promote more stakeholder consciousness and awareness around AI medical device development and usage.

Subject Keywords

Artificial Intelligence, Medical Device, Food and Drug Administration (FDA), Risk Mitigation, Life Cycle Processes

URI

https://hdl.handle.net/11511/117369

Collections

Graduate School of Informatics, Thesis