DIA 2025 Global Annual Meeting

In the rapidly evolving landscape of medical product development, the utilization of synthetic data and digital twins has gained increasing interest as a transformative approach to accelerate the development and approval of medical products. This session aims to explore the potential and innovative applications of these technologies in the realm of drug development, featuring speakers from health authorities, academia and industry. Dr. Khaled El Emam, Professor at the University of Ottawa, will present his latest research on generating synthetic data to augment clinical trials. Additionally, he will explore practical applications of synthetic data generation using real-world data. Dr. Arman Sabbaghi, Associate Professor at Purdue University, with prior experience at Unlearn, will present AI-Generated Digital Twins to deliver more efficient clinical trials. He will discuss the innovative statistical methodologies and novel trial designs that combine historical data, artificial intelligence (AI), and randomization to deliver smaller, faster, and more powerful RCTs that are built with regulatory guidance in mind. Dr. Ye Li, Mathematical Statistician, CDER, FDA will present statistical challenges leveraging machine learning in clinical trials. Several key factors that may influence the application of ML and AI algorithms in analyzing efficacy data will be discussed, including estimand and type I error rate control, as well as the importance of interpretability and reproducibility of ML/AI models to ensure robust and reliable results. Furthermore, she will address corresponding considerations related to these factors when using Digital Twins.