The Results section presents reusable frameworks and tools developed during the design phase of our study, enabling the registration and monitoring of 26 AI models and 2 ambient AI pilots. A retrospective analysis of 33 clinical nomograms demonstrates how these frameworks support our Responsible Artificial Intelligence (RAI) governance. The AI Task Force identified key challenges, including securing high-quality data and AI talent, leading to a Model Inventory of 87 active projects across research, clinical, and operational domains. The AI Governance Committee developed the iLEAP model, ensuring legal, ethical, and performance assessment throughout AI development. The AIGC manages an AI model portfolio, tracking AI’s lifecycle stages in the Model Registry, and evaluating performance using key metrics post-deployment. Case studies illustrate our governance in action for both acquired and internally built models, emphasizing ongoing monitoring and the dynamic nature of AI demand, which has risen by 63% in 2024 compared to the previous year.
Source link
Share
Read more