Deploying AI to Better Suss Out HER2 Status

Findings from a multinational study suggest that adding an AI-integrated platform to training masterclasses for pathologists boosts their accuracy in classifying HER2-stained breast cancer samples, particularly when it comes to pinpointing “low” and “ultralow” categories. Incorporating AI assistance also appears to reduce the risk of HER2-null mislabeling, which would otherwise deny patients access to potentially beneficial targeted therapies.