Using 3D Imaging and AI to Improve Colon Cancer Care

This project introduces a new approach to colon cancer diagnostics by combining 3D tissue imaging, artificial intelligence, and spatial protein analysis. By analyzing entire tumor samples in three dimensions, it captures critical features—such as tumor heterogeneity, architecture, immune infiltration, and invasive behavior—that cannot be comprehensibly assessed by conventional pathology.

The aim is to improve therapy decisions for stage II colon cancer patients, where current methods cannot reliably predict who benefits from additional treatment. The project will deliver a clinically compatible AI-based decision support tool that enables more accurate risk stratification, helping to avoid over- or under-treatment.

The consortium brings together leading expertise in large-scale 3D imaging and tissue clearing (Ali Ertürk, Helmholtz Munich), population-based cancer research (Michael Hoffmeister, DKFZ), and AI-driven risk prediction (Titus Brinker, DKFZ), ensuring strong clinical relevance and translational impact.

In the long term, this technology will generalize to other tumor types and offers strong translation and start-up potential, providing the basis for scalable AI-driven 3D pathology solutions for routine clinical use.