How AI can revolutionize breast cancer screening
A recent large-scale trial in Sweden involving over 80,000 women found that using artificial intelligence (AI) in breast cancer screening is safe and can reduce radiologists' workload by nearly 50%. The study compared AI-supported screening to standard care, revealing that AI was as effective as two radiologists working together, and didn't increase false positives. The use of AI led to 36,886 fewer screen readings.
AI reduced 44% of radiologists' screen-reading workload
Unlike previous studies, this trial directly compared AI-supported mammography screening with standard care. Half of the scans were analyzed by two radiologists, while the remaining were assessed by AI-supported screening followed by further evaluation by one or two radiologists. Both groups had a false-positive rate of 1.5%. The results showed that AI brought a 44% reduction in radiologists' screen-reading workload.
AI-based screenings could allow radiologists to focus on advanced diagnostics
The findings of this study could impact breast cancer screening programs, radiologists' workloads, and patient outcomes. Implementing AI in screenings could allow radiologists to focus on advanced diagnostics and reduce patient wait times. However, more research is needed to confirm the effectiveness of AI in mammography screening, including its ability to detect interval cancers, which are detected between screenings.