Prospective deployment of AI-based risk stratification to enable expedited mammography workflow in a safety-net setting
AI-powered breast cancer screening dramatically shortened wait times and caught 26 times more cancers in high-risk patients at an urban safety-net hospital.
This prospective HIPAA-compliant study deployed the Mirai AI risk model in real-time within screening mammography at a UCSF urban safety-net facility (n=4145 mammograms), flagging the top 10% of risk for same-day expedited interpretation and diagnostic evaluation. Expedited patients had dramatically shorter timelines to results (99.1% reduction) and biopsy (87.2% reduction), and a cancer detection rate of 60/1000 compared to 2.3/1000 in non-high-risk participants, validating AI-driven risk stratification as an implementable disparity-reduction tool.
What the study was
- Study design
- Prospective controlled study
- Population
- Women undergoing screening mammography at urban safety-net facility (UCSF), predominantly underserved/minority population
- Sample size
- 4145
- Category
- Early Detection
- Maturity
- Validated
- Journal
- NPJ Digital Medicine
Why it surfaced
Prospective real-world deployment of AI mammography risk stratification at safety-net facility demonstrating dramatic workflow improvements (99% time reduction to biopsy) and 26x higher cancer detection rate in high-risk group vs background. NPJ Digital Medicine (high-impact). Directly addresses screening disparities in underserved populations.
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.