Researchers at UC San Francisco and UC Berkeley have developed an artificial intelligence tool that slashes wait times for breast cancer diagnostics from weeks to hours. By using the Mirai AI model to identify high-risk patients from screening mammograms, clinicians at Zuckerberg San Francisco General Hospital successfully expedited care, reducing biopsy wait times from over two months to fewer than 10 days, according to a study published May 19 in Nature Digital Medicine.
How does AI change the mammogram screening process?
The Mirai model acts as a triage system rather than a diagnostic replacement for radiologists. According to Dr. Maggie Chung, the study’s lead author, the AI analyzes screening mammograms for subtle patterns associated with cancer risk that a physician might not immediately prioritize. By flagging the 12.7% of patients at the highest risk, the system allows hospitals to bypass standard administrative delays. Instead of waiting for a follow-up appointment, these patients can receive same-day diagnostic imaging and, in some cases, same-day biopsies.
What are the primary benefits for patients?
The most immediate impact is the reduction of diagnostic uncertainty. Patients previously faced weeks of anxiety while waiting for results; the UCSF-UC Berkeley study reduced this window to approximately one hour. For patients who require biopsies, the acceleration is even more significant. Data from the study shows that wait times for biopsy procedures dropped from more than 60 days to under 10 days. This shift toward personalized, accelerated care ensures that interventions occur at the most effective time, rather than following a one-size-fits-all scheduling system.
How does this technology fit into clinical practice?
The AI serves as a collaborative partner for physicians, according to Dr. Yala. It does not issue final diagnoses, but it creates a data-driven workflow that helps clinicians manage high volumes of screenings more efficiently. In the study, researchers tested the algorithm on more than 114,000 archival mammograms to ensure the triage system would not overwhelm clinic resources. By balancing sensitivity with operational capacity, the team demonstrated that AI can integrate into existing hospital infrastructure without disrupting standard care, provided the system is calibrated to identify the most urgent cases.
Pro Tips for Understanding AI in Radiology
- AI is a tool, not a doctor: Always remember that radiologists make the final diagnosis. AI is used to organize the workflow.
- Risk-based screening: Ask your provider if your facility uses risk-assessment tools that prioritize patients based on individual breast density or personal history.
- Stay informed: Technology like Mirai is rapidly evolving; check the full study in Nature Digital Medicine for technical details on how the model functions.
Frequently Asked Questions
Does the AI replace my radiologist?
No. According to the UCSF study, the AI acts as a triage tool that identifies high-risk patients so doctors can prioritize them for faster evaluation.
How much time does this save?
The study found that diagnostic evaluation wait times dropped from several weeks to about one hour, while biopsy wait times decreased from over two months to fewer than 10 days.
Is this technology available everywhere?
Currently, the study focused on implementations at Zuckerberg San Francisco General Hospital. While promising, it represents an emerging practice in computational precision health that is still moving toward wider adoption.
Have you or a loved one experienced long waits for diagnostic results? Share your thoughts in the comments below or subscribe to our health technology newsletter for the latest updates on medical AI.
Keep reading