Landmark Study Suggests AI Could Revolutionize Breast Cancer Screening - Video

A new study from Sweden suggests that artificial intelligence (AI) can significantly improve breast cancer screening by detecting cancers earlier and reducing diagnoses after routine screening. Published in The Lancet, the research involved nearly 100,000 women who underwent mammography between April 2021 and December 2022. This represents the first large randomized trial assessing AI’s performance in real-world screening programs.

Participants were randomly divided into two groups. In the standard screening group, mammograms were read by two radiologists. In the AI-supported group, an AI system initially evaluated all scans. Low-risk cases were reviewed by one radiologist, while higher-risk cases were assessed by two radiologists, with AI highlighting suspicious areas.

The findings were notable. Fewer women in the AI group were diagnosed with breast cancer in the years following screening—1.55 cases per 1,000 women versus 1.76 per 1,000 in the standard group, representing a 12% reduction. AI-supported screening also detected a higher proportion of cancers during the screening itself (81% compared with 74%) and reduced aggressive cancer subtypes by 27%.

Dr. Kristina Lang, lead author and researcher at Lund University, said the study demonstrates that AI can improve early detection while easing the workload for radiologists. However, she emphasized that AI should support clinicians rather than replace them, and its use must be carefully monitored to ensure accuracy and safety.

Experts welcomed the results but noted that additional studies are needed in different countries and healthcare systems to confirm these findings and evaluate long-term effectiveness.

Overall, the study highlights AI’s potential to enhance breast cancer screening, enabling earlier diagnosis, reducing late-stage disease, and potentially improving patient outcomes. With careful integration into clinical practice, AI-assisted mammography may become an important tool in the fight against breast cancer.

REFERENCE: Gommers, Jessie et al.; Interval cancer, sensitivity, and specificity comparing AI-supported mammography screening with standard double reading without AI in the MASAI study: a randomised, controlled, non-inferiority, single-blinded, population-based, screening-accuracy trial; The Lancet, Volume 407, Issue 10527, 505 – 514; doi: 10.1016/S0140-6736(25)02464-X