Session Description:

From the rapid development and dissemination of AI systems within the general discourse, the discussion within the scientific community now centers on in-house AI tools that promote data security, confidentiality, and domain specificity while also retaining the user-friendliness and sophistication of general purpose AI systems. However, the development of such AI tools, as is necessary with any computational model, requires that a rigorous and empirically-derived process of evaluation and validation be implemented. This session discusses the processes of evaluating AI tools—with a specific focus on clinical prediction models—using both quantitative and qualitative methods. Recommendations for clinical practice, implementation of these AI tools, and future research are provided.

Presented by:

Tony Mangino, PhD
Assistant Professor of Biostatistics

Register »