Eireirnews
All stories
clinical

GEMINI Study: AI Boosts UK Breast Cancer Detection by 10.4% While Cutting Workload by a Third

The GEMINI study, published in Nature Cancer, found that integrating AI into UK breast cancer screening increased cancer detection by 10.4%, reduced recall rates, cut workload by up to 31%, and slashed cancer notification time from 14 days to 3 days.

Source: The ASCO Post
breast-cancermammographyAI-screeningNHSNature-CancerGEMINI

The GEMINI study, published in Nature Cancer, has delivered the most comprehensive evidence yet for AI integration in population-based breast cancer screening. Evaluating the Mammography Intelligent Assessment (Mia) v.3 tool alongside routine human review of 10,889 mammograms in the UK NHS, the results show AI increased cancer detection by 10.4% — one additional cancer found per 1,000 patients screened.

The operational gains are equally striking. Recall rates decreased by 0.8%, meaning fewer women were subjected to unnecessary follow-up anxiety. Workload reductions reached up to 31%, and total cost savings hit 36% compared to standard double-reading processes. Perhaps most impactful for patients, cancer notification time dropped from 14 days to just 3 days.

Lead researcher Dr. Clarisse Florence de Vries summarized the findings: 'We found optimal ways to detect breast cancer, quicker, and more accurately. We also found ways to reduce the number of women having to return for unnecessary tests.' The team plans to expand these findings through the upcoming EDITH trial across the broader UK.

The study arrives as the UK faces a 29% shortage of clinical radiologists. With breast cancer affecting one woman every 10 minutes in the UK, AI's ability to maintain or improve detection quality while reducing radiologist burden addresses both a clinical and workforce challenge simultaneously.

This story was produced by an automated system. Always verify critical information with the original source.

Last updated: Monday, March 23, 2026

Related stories

clinical

AI Outperformed Physicians in Hospital Patient AVS Tasks, Raising the Bar for Clinical Documentation Tools

Medscape reports that AI beat physicians on AVS tasks for hospital patients, underscoring how administrative and documentation work may be easier for machines to standardize than for overextended clinicians. The finding does not eliminate human oversight, but it does strengthen the case for AI in workflow support roles.

clinical

FDA Clears New AI Sepsis Tool as Hospitals Keep Pushing for Earlier Intervention

The FDA has cleared an AI sepsis tool that its developer says can detect infection hours earlier than clinicians. The approval adds momentum to one of the most closely watched categories in hospital AI: systems that promise to identify deterioration before it becomes irreversible.

clinical

AI Sepsis Tools Are Moving From Promise to Proof, but the Real Test Is in Workflow

AI sepsis tools are attracting renewed attention as they gain traction in hospitals and regulators. The challenge now is not technical novelty but whether these systems can improve outcomes without overwhelming clinicians with noise.