AI-based estimation of tumor cell percentage in NSCLC, CRC and breast cancer: agreement analysis and feature overview from digital pathology algorithms

Background & Objectives

Estimation of tumor cell percentage (TC%) is a mandatory pre-analytical step before molecular testing in non-small cell lung cancer (NSCLC) yet remains poorly reproducible and imprecise when performed manually by pathologists. While tumor recognition relies on pathologist expertise, TC% quantification is a mathematical task for which the human eye is poorly suited.

Conference

European Congress of Pathology 2025