the U.S. Food and Drug Administration (FDA) on track to clear its 1,000th clinical artificial intelligence (AI) algorithm by the end of 2024, FDA Commissioner Robert Califf, MD, He emphasized the risks of relying solely on pre-market approvals, noting the potential for AI algorithms to deviate, or "drift," from their original, FDA-cleared functions over time.
This gap in quality assurance to check for AI drift poses significant risks, particularly for critical applications like stroke prediction, where changing patient demographics and data can undermine the accuracy of AI-driven predictions.
A significant obstacle to effective AI oversight, is the fragmented nature of healthcare data. While individual patient records can often be accessed across systems, the aggregate data needed to evaluate AI performance is siloed, limiting the ability to assess sensitivity, specificity and overall reliability.
A fundamental shift in how health systems manage AI, urging the creation of collaborative, data-driven ecosystems that prioritize transparency and accountability. The importance of equipping health systems with the technical expertise and resources needed to monitor AI performance over time.
【MORE】