AI use in breast radiology increased cancer detection without higher recall rates, according to a 100,000+ case multicenter study.
Key Details
- 1Study conducted across four sites and included nine experienced breast radiologists.
- 2Over 100,000 digital breast tomosynthesis (DBT) exams analyzed.
- 3Results compared pre- and post-AI implementation: 54,000 exams before AI (339 true positives) vs. nearly 49,000 with AI (369 true positives).
- 4AI improved invasive and dense breast cancer detection and lowered average diagnosis stage.
- 5No increase in recall rates was observed with AI usage.
- 6The AI tool was provided by iCAD and the DBT system by GE HealthCare.
Why It Matters
Demonstrates AI's practical benefits in increasing breast cancer detection while maintaining recall rates, addressing workload and burnout concerns. This adds impactful real-world data supporting AI adoption in radiology practices.
Source
Radiology Business
Related News
•Radiology Business
NYC Health + Hospitals CEO Considers AI to Replace Radiologists
NYC Health + Hospitals CEO suggests AI could partially replace radiologists, pending regulatory approval.
•AuntMinnie
AI Models Reveal Racial Disparities in Breast Cancer Patterns
Machine learning models reveal significant racial disparities and key predictors in breast cancer incidence across diverse groups.
•Radiology Business
Real-World Trial: AI Tool Cuts Radiologist Workload in Mammography
A prospective study shows AI can safely reduce radiologists’ workload and increase cancer detection in breast cancer screening.