
Breast Imaging in Focus: Key Takeaways from the Recent Lancet Mammography Study on AI and Interval Breast Cancer Rates
In the debut of her new “Breast Imaging in Focus” series, Manisha Bahl, M.D., discusses the recently published Lancet study on AI, mammography and interval breast cancer, and shares her perspective on the potential impact of these findings for breast radiologists.
Maintaining or improving breast cancer detection amid increasing imaging workloads and radiologist shortages is a worldwide challenge for breast cancer screening programs.
Accordingly, there has been a fair amount of buzz in the breast imaging community about the recently published
Researchers found that adjunctive AI facilitated a significantly higher cancer detection rate and a non-inferior interval breast cancer rate with a 44 percent reduction in reading workload in comparison to standard double reading of mammograms in Sweden without AI assistance.
From my perspective, the most important findings are the combination of higher cancer detection, non-inferior interval cancer rates and dramatically reduced radiologist workload, The fact that interval cancers were not increased and that sensitivity was actually higher addresses one of the main concerns about using AI as a gatekeeper in screening. Equally important is that these gains were achieved without increasing recall rates, which is critical for patient experience and downstream costs.
The Lancet study demonstrates that adjunctive AI can help maintain quality diagnostic assessment and facilitate appropriate triage, allowing breast radiologists to spend more time focusing on cases requiring a higher level of evaluation.
For breast imaging radiologists, these findings suggest a future in which AI can function as a true partner. AI-supported screening could allow us to focus more attention on complex exams and reduce fatigue associated with high volume screening at a systems level. This approach may help preserve high quality population-based screening in the face of growing demand and limited radiology resources without sacrificing diagnostic performance.
(Editor’s note: For related content, see “















