Google researchers have published two companion studies in Nature Cancer examining how AI could improve breast cancer screening workflows in the UK's National Health Service. The studies assessed an AI system's standalone performance and its integration into existing double-read workflows, addressing critical radiologist shortages projected to reach 40% by 2028.

In the first study, the AI system analyzed mammograms from 115,973 women across five NHS screening services, achieving significantly higher sensitivity than human readers without compromising specificity. The system detected 25% of interval cancers missed in the original double-read workflow and performed exceptionally well for first-time screenings, reducing false positives while increasing detection rates from 7.54 to 9.33 per 1,000 women.

A prospective deployment at 12 London screening sites demonstrated successful technical integration, with AI processing cases in a median of 17.7 minutes compared to over two days for human readers. The second study involved 22 accredited mammography readers arbitrating 8,732 cases using both standard and AI-enabled workflows. These findings suggest AI could help sustain breast cancer screening programs while maintaining rigorous standards, though researchers emphasize that additional work is needed to prove effectiveness in prospective clinical practice.