New scientific publication by University of Navarra

Neoadjuvant therapy (NAT) is the primary treatment for reducing breast cancer tumor size before surgery. However, predicting how tumors respond varies based on patient factors. To enhance personalized care, we’re developing a machine learning (ML) model. It combines imaging and molecular data to forecast NAT response. Our model uses two stages: first, a non-invasive MRI model, and second, a biomarker-enriched model. We integrate a framework called Conformal Prediction (CP) to identify patients needing further testing, reducing unnecessary biopsies. Testing on real clinical data shows promising results for our predictive tool.

Download the last publication about the project

An Uncertainty-Aware Sequential Approach for Predicting Response to Neoadjuvant Therapy in Breast Cancer / Alberto Garcia-Galindo, Marcos Lopez-De-Castro, Ruben Armananzas