AI Boosts Breast Cancer Detection in World-First Trial, Reducing Aggressive Cases

STOCKHOLM — Artificial intelligence (AI) has helped doctors detect significantly more cases of breast cancer during routine screenings while reducing the emergence of aggressive tumors between check-ups, according to the final results of a landmark randomized controlled trial published Friday in The Lancet.

The study, the first of its kind to measure the real-world clinical impact of AI in a national screening program, found that AI-supported screening identified 9% more cancer cases than standard methods. Perhaps more crucially for patient outcomes, it led to a 12% reduction in “interval cancers”—cancers that appear in the two-year gap between regular screening appointments and are often the most difficult to treat.

Led by researchers at Lund University in Sweden, the MASAI (Mammography Screening with Artificial Intelligence) trial followed more than 100,000 women over two years. The findings suggest that AI could not only improve survival rates through earlier detection but also provide a critical solution to the global shortage of specialist radiologists.

A New Standard for Screening?

In most European countries, including Sweden, the “gold standard” for mammography involves two radiologists independently reading every scan to minimize human error. The MASAI trial challenged this labor-intensive model.

Participants were divided into two groups:

1. AI-Supported Group: A single radiologist used an AI system (Transpara) to triage and analyze scans. High-risk scans were flagged for human review, while the AI assisted in identifying suspicious tissue.

2. Control Group: Two radiologists independently read each scan without AI assistance.

The results were striking. The AI group detected 338 cancers, compared to 262 in the standard group. Despite the increased detection, the rate of false positives—where a woman is called back for a biopsy only to find no cancer—remained nearly identical between the groups (1.5% vs 1.4%).

“Our study is the first randomized controlled trial investigating the use of AI in breast cancer screening,” said Dr. Kristina Lång, lead author and radiologist at Lund University. “The findings show that AI-supported screening improves the early detection of clinically relevant breast cancers, which led to fewer aggressive or advanced cancers diagnosed in between screenings.”

Catching “Hidden” Threats

One of the most significant findings involves the reduction of interval cancers. These tumors are often more biologically aggressive and are missed by standard screening because they are either too small or obscured by dense breast tissue.

The trial recorded 1.55 interval cancers per 1,000 women in the AI group, compared to 1.76 in the control group. Among these, the AI group saw 27% fewer aggressive subtypes (non-luminal A cancers), which typically carry a poorer prognosis.

Implications for a Strained Workforce

Beyond clinical outcomes, the trial addressed the practicalities of a healthcare system under pressure. Interim data from the same study showed that AI-supported screening reduced the screen-reading workload for radiologists by 44%.

With many countries facing a “ticking time bomb” of radiologist retirements and rising screening demands, the ability for one doctor to perform the work of two—without losing accuracy—is a major development.

“Widely rolling out AI-supported mammography could help reduce workload pressures,” Dr. Lång noted, while emphasizing that human oversight remains essential.

Expert Perspectives and Caveats

While the results are being hailed as a “paradigm shift,” independent experts urge a measured approach to implementation.

Professor Stephen Duffy, an emeritus professor of cancer screening at Queen Mary University of London, noted that while the safety and efficacy are clear, the 12% reduction in interval cancers, though promising, did not reach full statistical significance in this specific timeframe. “We need longer follow-up to see if these trends translate into a definitive reduction in breast cancer mortality,” Duffy said.

In France, where AI use remains limited, Jean-Philippe Masson, head of the French National Federation of Radiologists, expressed concerns about the high costs of these systems and the potential for “overdiagnosis”—the detection of tiny, slow-growing lesions that might never have caused harm if left alone. “The radiologist’s experience must still correct the AI’s diagnosis,” Masson warned.

What This Means for Patients

For the average woman attending a screening, the integration of AI likely means a more thorough “second pair of eyes” reviewing her scans.

• Accuracy: AI performs consistently across different ages and breast densities.

• Earlier Intervention: Detection of smaller, lymph-node-negative cancers often allows for less invasive treatments, such as lumpectomies instead of mastectomies.

• No Added Stress: Because the false positive rate did not increase, patients are not at a higher risk of unnecessary anxiety or biopsies.

The researchers conclude that while AI is ready for “controlled implementation,” it should be accompanied by rigorous quality monitoring to ensure the algorithms perform as expected across diverse populations.

Statistical Overview: MASAI Trial Results

Medical Disclaimer: This article is for informational purposes only and should not be considered medical advice. Always consult with qualified healthcare professionals before making any health-related decisions or changes to your treatment plan. The information presented here is based on current research and expert opinions, which may evolve as new evidence emerges.

References

• Journal Publication: Gommers, J., Lång, K., et al. (2026). “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. DOI: 10.1016/S0140-6736(25)02464-X.