Title: High-Fidelity Synthetic Data Applications for Data Augmentation
Speaker: Puja Myles Director, Clinical Practice Research Datalink (CPRD) Safety and Surveillance group. UK Medicines and Healthcare products Regulatory Agency
Abstract: This presentation will provide an overview of the UK Medicines and Healthcare products Regulatory Agency’s (MHRA) work on high-fidelity synthetic data applications with a focus on data augmentation to address biases due to underrepresentation of population subgroups and boosting small sample sizes in the context of clinical trials. The presentation will also include a brief overview of the MHRA’s approach to synthetic data generation and evaluation as well as our emerging thinking on good practice in this area.
Title: How the HDL is Enabling Health Data Research with AI
Speaker: Steffen Hess Head of Health Data Lab, The Federal Institute for Drugs and Medical Devices (Bundesinstitut für Arzneimittel und Medizinprodukte, BfArM)
Abstract: The Health Data Library (HDL) manages the claims data of 74 million individuals in Germany, characterized by its sensitive nature and high level of detail that results in numerous unique datasets. To address privacy concerns, HDL implements a range of privacy-enhancing measures. These include an AI-powered sandbox and the generation of AI-based synthetic data, all within a secure processing environment. The focus is on striking a balance between safeguarding data privacy and maintaining data usability. HDL is committed to ongoing evaluations of both the security and practical utility of these measures to ensure optimal protection and functionality.