What happened
The UK government announced a trial across 30 sites involving nearly 700,000 women already attending NHS breast screening. Backed by £11 million, EDITH is designed to test AI-assisted mammogram reading in routine care.
Why it matters
This is a health-system implementation test, not simply an algorithm benchmark. The result has implications for diagnostic capacity, radiologist workload and the standard of evidence required before AI changes national screening workflows.
What to watch
Cancer-detection and recall rates, interval cancers, performance across demographic groups, radiologist oversight, and whether operational gains persist outside the trial.
