A comprehensive evaluation of Artificial Intelligence in the UK breast cancer screening program has revealed that AI can increase cancer detection rates by 10.4% while reducing healthcare worker workload by over 30%. Published today in Nature Cancer, the study was conducted by a collaboration of scientists and clinicians from the University of Aberdeen, NHS Grampian, and Kheiron Medical Technologies, now part of DeepHealth Inc. The research focused on the GEMINI project, which utilized the Mia AI software tool to support the routine screening of 10,889 women in NHS Grampian. Under the current UK system, mammograms are typically reviewed by two human radiologists to minimize missed diagnoses. However, this process remains imperfect, with approximately 20% of cancers missed. Furthermore, the high rate of false positives leads to unnecessary recalls and biopsies; for every five women called back for further testing, only one is diagnosed with cancer, causing significant patient anxiety and straining medical resources. The GEMINI trial tested seventeen different scenarios for integrating AI into the screening workflow. The most effective configuration involved using AI as both a second reader, substituting for one human radiologist in normal cases, and as an additional safeguard. This hybrid approach successfully increased the detection of invasive, high-grade cancers while significantly lowering the number of unnecessary recalls. Beyond detection rates, the study highlighted critical improvements in patient outcomes and operational efficiency. The integration of AI reduced the average time to notify affected women from 14 days to just three days. Early detection of aggressive cancers is vital as it allows for earlier treatment and a higher likelihood of success. Additionally, reducing the volume of unnecessary biopsies and follow-up tests alleviates patient stress and conserves healthcare resources. Dr. Clarisse de Vries, lead author and lecturer in Data Science at the University of Glasgow, emphasized the significance of these findings. She noted that while the UK National Screening Committee has previously hesitated to recommend AI due to insufficient evidence, this study provides high-quality data supporting its safety and efficacy. The research demonstrates that AI can be tailored to local healthcare needs to enhance service delivery. Niccolo Stefani of DeepHealth stated that the study proves AI can reimagine care delivery by improving accuracy and enabling proactive, scalable solutions. Professor Lesley Anderson from the University of Aberdeen praised the trial's unique design, which simulated real-world usage to help policymakers understand operational integration. The team found that AI effectively augments radiologist practice, potentially reducing burnout among clinicians facing shortages and high workloads. While further research is needed to fully quantify long-term benefits and potential harms, these findings address key evidence gaps identified by national health bodies. The results will directly inform the upcoming EDITH trial, a larger study designed to evaluate AI in breast screening across multiple sites throughout the UK. Led jointly by the University of Aberdeen, NHS Grampian, and the University of Glasgow, this next phase aims to build trust and accelerate the adoption of clinical AI. Professor Mike Lewis of the NIHR noted that generating robust evidence on the safe and effective use of AI is essential for delivering tangible improvements across the health system.