Artificial intelligence-supported mammography screening reduced interval cancer rates by 12% and improved detection sensitivity compared with standard double reading, according to results from the Mammography Screening With Artificial Intelligence (MASAI) study published in The Lancet. Interval cancers are breast cancers diagnosed between screening rounds that were not detected at initial screening. The AI-supported approach also reduced screen reading workload by 44%, suggesting it may be considered for implementation in clinical practice, particularly amid radiologist workforce shortages.

Study Methodology

Between April 12, 2021, and December 7, 2022, researchers randomly assigned 105,934 women in a 1:1 ratio to either AI-supported mammography screening (the intervention group) or standard double reading without AI (the control group). Researchers used AI to triage examinations to single or double reading by radiologists and for detection support. Median age was 54 years in both groups.

Results

Interval cancer rates were 1.55 per 1,000 patients in the AI-supported group compared with 1.76 per 1,000 patients in the control group, a non-inferior proportion ratio of 0.88.

Patients in the intervention group had fewer interval cancers that were invasive (75 vs 89, representing 16% fewer). There were 10 fewer T2+ interval cancers, which are tumors measuring greater than 20mm (38 vs 48), and 16 fewer non-luminal A cancers (43 vs 59, representing 27% fewer) in the AI group compared with the control group.

Sensitivity was higher in the AI-supported group (80.5%) compared with the control group (73.8%), an effect consistent across age and breast density, and for invasive cancer but not for in-situ cancer. Specificity was 98.5% for both groups.

Clinical Significance

"The MASAI trial showed consistently more favorable outcomes with AI-supported mammography screening compared with standard double reading without AI, including the primary outcome of interval cancer rate, showing non-inferiority, and fewer interval cancers with unfavorable characteristics. Further analyses of subsequent screening rounds and cost-effectiveness will clarify the long-term balance of benefits and harms, and could provide a strong rationale for implementing AI in population-based mammography screening programs, particularly in the context of [radiologist] workforce shortages," wrote lead study author Kristina Lång, PhD, of Lund University in Sweden, and colleagues.

Funding for this study was provided by the Swedish Cancer Society, the Confederation of Regional Cancer Centres, and Swedish governmental funding for clinical research.

Full disclosures for all authors are available with the full study.

Source: The Lancet