PRECISE-CRC: PREvention through Causal Inference and Stratified Embeddings in ColoRectal Cancer
Hector Fellow Bernhard Schölkopf
Hector RCD Awardee Carolin Schneider
PRECISE CRC is developing an embedding-based system that transforms heterogeneous lifestyle and healthcare data from biobanks into structured patient summaries using a context-specific large language model (LLM), embeds these into latent vectors, and identifies true causal risk factors for colorectal cancer using advanced causal methods to identify true causal risk factors for colorectal cancer.









