AI as a Safety-Net Reader for Mammograms Classified as Normal or Benign in the French Screening Program.

Authors

Abstract

In the French national breast cancer screening program, second reading is performed only for mammograms interpreted as negative (BI-RADS 1-2) at first reading, representing a unique screening workflow in Europe. This retrospective study assessed whether artificial intelligence (AI) could identify a subgroup of negative screening mammograms that could safely bypass second reading. A total of 55,589 screening mammograms from 42,419 women aged 50-74 years (January 2015-December 2019) initially classified as BI-RADS 1-2 were analyzed. Second-reading outcomes were compared with those of a commercial AI system using a predefined binary threshold (≥ 5). Among these examinations, 183 of 55,589 (0.33%) were recalled at second reading, yielding 12 cancers (positive predictive value, 6.6%; cancer detection rate, 0.22 per 1,000 examinations). AI classified 42,606 of 55,589 (76.6%) examinations as low risk (≤ 4) and 12,983 of 55,589 (23.3%) as non-low risk (≥ 5). One cancer was detected in the AI-low group (1 of 55,589 [0.002%]) compared with 11 in the AI-nonlow group (11 of 55,589 [0.020%]; <i>P</i> < .001). Interval cancer rates were higher in the AI-non-low group than in the AI-low group (2.16 vs 0.47 per 1,000 examinations). These findings suggest that excluding AI-low examinations from second reading could reduce workload by approximately 77% while focusing radiologist review on higher-risk cases; prospective validation is needed. AI triage could potentially reduce second-reading workload by approximately 77% in the French breast cancer screening program, despite a small but measurable risk of missed cancers and the need for governance.

Tags

Topics