New AI model improves cancer diagnosis accuracy significantly

A new AI model, CHIEF, developed by researchers at Harvard Medical School, shows up to 36% improved effectiveness in diagnosing multiple cancers compared to existing methods. Trained on over 15 million pathology images, it achieved nearly 94% accuracy across 11 cancer types. The model's performance reached 99.43% in specific applications. Further validation in diverse clinical settings is needed before regulatory approval.