On April 3, 2025 at 11:00 a.m. the lecture of the Living Lab lecture series will take place. In this talk, Waldemar Hahn will talk about the Generation of Synthetic Tabular Data in the Medical Domain.

Access to medical data is critical for advancing healthcare, yet legal and privacy constraints make real-world datasets difficult to share. Synthetic tabular data offers a promising alternative by mimicking real distributions without exposing patient identities. In this talk, we provide a high-level overview of how generative models, such as Variational Autoencoders (VAEs) and Generative Adversarial Networks (GANs), originally developed for image generation, are adapted to the structure and challenges of tabular data. We then explore how the resulting synthetic data can be evaluated in terms of resemblance, utility, and privacy. Finally, we outline what a realistic pipeline for publishing medical synthetic tabular data might look like in practice.