Results from large-scale trial show AI technology cut interval cancer rates while maintaining screening accuracy across patient populations.
AI-supported mammography screening significantly reduced aggressive breast cancers and interval cancers in the first randomized controlled trial of breast AI technology, according to final results published in The Lancet.
The Mammography Screening with Artificial Intelligence (MASAI) trial, involving more than 105,000 women in Sweden, found that ScreenPoint Medical’s Transpara Detection technology contributed to a 12% reduction in interval cancer rates and a 27% decrease in aggressive cancers compared to standard double-reading protocols.
Interval cancers are breast cancers diagnosed between screening rounds or within two years after scheduled screening that were not detected during the original screening. These cancers are associated with higher mortality rates compared to screen-detected cancers.
“MASAI demonstrates that breast AI has crossed a critical threshold: Right now, it is shaping a healthier future for populations at scale, it is allowing healthcare providers to do their best and most efficient work, and most importantly, it is having a profound impact for women around the world,” says Pieter Kroese, chief executive officer of ScreenPoint Medical, in a release.
Key Clinical Outcomes
The trial showed several clinically relevant improvements in the AI-supported screening group compared to controls:
- 16% fewer invasive interval cancers overall
- 21% fewer large (T2+) interval cancers
- 27% fewer non-luminal A subtype interval cancers
- 6.7% higher sensitivity (80.5% vs 73.8%) at the same specificity level (98.5%)
The results were consistent across different age groups and breast density categories, according to the study.
Previous MASAI research published in 2023 and 2025 demonstrated that the AI technology increased cancer detection rates by 29% while reducing radiologist workload by 44% compared to standard double-reading practices.
Clinical Implications
The findings suggest AI-supported screening may help detect clinically important breast cancers earlier while reducing the number of cancers that develop between regular screenings.
“These findings indicate that AI-supported mammography screening can help find important breast cancers earlier and reduce the number of cancers that appear between regular screenings, particularly those that are more aggressive,” says Dr Kristina Lång, lead researcher at Lund University, in a release. “As these types of cancers typically result in poorer outcomes, earlier detection may make a meaningful difference for women’s health.”
The trial achieved these improvements without increasing false positive rates, maintaining the same specificity as traditional screening methods while improving overall sensitivity.
Technology Details
Transpara Detection provides radiologists with AI-assisted analysis for both 2D and 3D mammography. The technology is backed by more than 55 peer-reviewed publications and has been shown to help overcome variables including breast density, patient ethnicity, and radiologist experience levels.
The system is currently used by healthcare organizations and screening programs globally as a supplementary tool to help radiologists detect cancers earlier while reducing reading workload.
The MASAI trial represents the largest randomized controlled study of AI in breast cancer screening to date, providing evidence for the clinical benefits of AI-supported mammography in population-level screening programs, according to a release from ScreenPoint Medical.
ID 87487092 © Tyler Olson | Dreamstime.com
