From June 25th till June 27th 2025, Heidelberg University hosted the second edition of the Sorbonne-Heidelberg AI in Medicine workshop, organized under the 4EU+ European University Alliance. This year's workshop, focused on multimodal data analysis, reflecting the growing importance of integrating diverse data types—from structured clinical records to unstructured text and omics data—in medical research.

A key contribution came from the EMCL group specializing in causal modeling for medical treatment effect estimation. EMCL led a improved hands-on session compared to the first edition workshop in 2024, addressing the challenges of applying neural networks to infer causal relationships from observational data, offering practical insights into the complexity and potential of this approach.

Author: Marcus Buchwald

Date:  27/06/2025 - Heidelberg