Invasive Breast Cancers Missed by AI Screening of Mammograms

Authors: Ok Hee Woo, Sung Eun Song, Su Jin Choe, Minhye Kim, Kyu Ran Cho, Bo Kyoung Seo

Published: 2025-06-24

DOI: 10.1148/radiol.242408

Source: Full article


Abstract

Among artificial intelligence–missed invasive breast cancers on mammograms, the human epidermal growth factor receptor 2–enriched subtype had the lowest false-negative rate, and dense breasts, nonmammary zone locations, architectural distortions, and amorphous microcalcifications were reasons for misses.